Synthetic Approaches to Piperazine-Containing Drugs Approved by FDA in the Period of 2011–2023 ^†

Abstract

The piperazine moiety is often found in drugs or in bioactive molecules. This widespread presence is due to different possible roles depending on the position in the molecule and on the therapeutic class, but it also depends on the chemical reactivity of piperazine-based synthons, which facilitate its insertion into the molecule. In this paper, we take into consideration the piperazine-containing drugs approved by the Food and Drug Administration between January 2011 and June 2023, and the synthetic methodologies used to prepare the compounds in the discovery and process chemistry are reviewed.

1. Introduction

Piperazine is among the most frequently used heterocycles in biologically active compounds [1,2,3]. This moiety is useful for different reasons: for its impact on the physicochemical properties of the final molecule, for its structural and conformational characteristics and for its easy handling in synthetic chemistry. Indeed, many surveys can be found in the literature on the application of the piperazine ring in biologically active compounds within different research areas (see [4] and the references cited therein).

In a previous paper, we discussed the role of the piperazine ring, first analyzing the drugs approved by the Food and Drug Administration (FDA) since 2017 that showed such a moiety, and then looking at biologically active piperazine derivatives for specific therapeutic areas [4]. As described there, the piperazine moiety is mainly used as a basic and hydrophilic group to optimize the pharmacokinetic properties of the final molecule or as a scaffold to arrange pharmacophoric groups in the proper position in the interaction with the target macromolecules. Our long-lasting interest in this field prompted us to also revise the synthetic procedures that have been mostly used in medicinal and process chemistry to prepare piperazine-containing drugs. It must be noticed that, owing to the popularity of this moiety, many useful synthons are commercially available: either (a) N-acyl or N-aryl piperazines decorated with protecting groups and/or with functional groups useful for further expansion of the molecule, or (b) piperazines carrying substituents such as phenyl, methyl or carboxylic acid, among others, on the ring C-atom.

The synthetic procedures developed to build the piperazine ring or to insert substituents have been the topic of some reviews [5,6,7,8]; such methods allow obtaining piperazine derivatives with a high degree of substitution on the ring. However, the structural complexity of the piperazine moiety in biologically active molecules varies considerably. We analyzed the structures of FDA-approved drugs from the period from January 2011 to June 2023 (

Table 1. New small molecules approved by FDA between January 2011 and June 2023, containing a piperazine ring with substituents only on the N-atoms.

Compound Number	Name, Year of Approval, Mechanism of Action and Therapeutic Indication	Structure
1	Palbociclib (2015): Cyclin-Dependent Kinase 4/6 inhibitor (treatment of metastatic breast cancer)
2	Ribociclib (2017): Cyclin-Dependent Kinase 4/6 inhibitor (treatment of metastatic breast cancer)
3	Vortioxetine (2013): serotoninergic modulator (treatment of major depressive disorder)
4	Avapritinib (2020): platelet-derived growth factor receptor alpha inhibitor (treatment of Gastrointestinal Stromal Tumor)
5	Letermovir (2017): cytomegalovirus DNA terminase inhibitor (to prevent infection in bone marrow transplant)
6	Trilaciclib (2021): Cyclin-Dependent Kinase 4/6 inhibitor (mitigation of chemotherapy-induced myelosuppression in small cell lung cancer)
7	Infigratinib (2021): fibroblast growth factor receptor inhibitor (treatment of cholangiocarcinoma)
8	Entrectinib (2019): ALK, ROS1 and Trk kinase inhibitor (treatment of metastatic non-small cell lung cancer)
9	Avatrombopag (2018): thrombopoietin receptor agonist (treatment of thrombocytopenia)
10 11	Netupitant (2014) and Fosnetupitant (2018): NK1 receptor antagonists (treatment of nausea and vomiting in patients undergoing cancer chemotherapy, in combination with palosetron)
12	Venetoclax (2016): Bcl-2 blocker (treatment of chronic lymphocytic leukemia in patients with a specific chromosomal abnormality)
13	Brexpiprazole (2015): (treatment of schizophrenia and major depressive disorder)
14	Vilazodone (2011): serotoninergic modulator (treatment of major depressive disorder)
15	Flibanserin (2015): 5-HT1A agonist (treatment of acquired, generalized hypoactive sexual desire disorder in premenopausal women)
16	Aripiprazole Lauroxil (2015): long-acting antipsychotic (treatment of schizophrenia)
17	Cariprazine (2015): D2/D3 receptors partial agonist (treatment of schizophrenia and bipolar disorder in adults)
18	Bosutinib (2012): Bcr-Abl tyrosine-kinase inhibitor (treatment of chronic myelogenous leukemia)
19	Ponatinib (2012): Bcr-Abl tyrosine-kinase inhibitor (treatment of chronic myeloid leukemia and Philadelphia chromosome positive acute lymphoblastic leukemia)
20	Nintedanib (2014): receptor and non-receptor tyrosine kinase inhibitor (treatment of idiopathic pulmonary fibrosis)
21	Abemaciclib (2017): Cyclin-Dependent Kinase 4/6 inhibitor (treatment of metastatic breast cancer)
22	Gilteritinib (2018): FMS-like tyrosine kinase 3 inhibitor (treatment of Acute Myeloid Leukemia)
23	Brigatinib (2017): anaplastic lymphoma kinase/epidermal growth factor receptor inhibitor (treatment of non-small cell lung cancer)
24	Maralixibat (2021): ileal bile acid transporter inhibitor (treatment of cholestatic pruritus associated with Alagille syndrome)
25	Mitapivat (2022): pyruvate kinase activator (treatment of hemolytic anemia in pyruvate kinase deficiency)
26	Zavegepant (2023): calcitonin gene-related peptide receptor antagonist (treatment of migraines)
27	Olaparib (2014): poly ADP ribose polymerase inhibitor (treatment of advanced ovarian cancer)
28	Fostemsavir (2020): HIV attachment inhibitor (treatment of HIV infection)

and

Table 2. New small molecules approved by FDA between January 2011 and June 2023, containing a piperazine moiety with substituents in the ring C-atoms or inserted into a polycyclic structure.

Compound Number	Name, Year of Approval, Mechanism of Action and Therapeutic Indication	Structure
29	Selpercatinib (2020): Rearranged during Transfection (RET) inhibitor (treatment of lung and thyroid cancers)
30	Risdiplam (2020): Survival Motor Neuron-2 RNA splicing modifier (treatment of spinal muscular atrophy)
31	Sotorasib (2021): KRASG12C inhibitor (treatment of non-small cell lung cancer)
32	Adagrasib (2022): KRASG12C inhibitor (treatment of locally advanced or metastatic non-small cell lung cancer)
33	Fezolinetant (2023): NK3 receptor antagonist (treatment of moderate to severe hot flashes caused by menopause)
34	Lumateperone (2019): 5-HT2A/D2 antagonist (schizophrenia)
35	Fosdenopterin (2021): molybdenum cofactor precursor (to reduce the risk of mortality in patients with molybdenum cofactor deficiency Type A)
36	Lurbinectedin (2020): DNA minor groove binder (treatment of metastatic small cell lung cancer)
37	Trabectedin (2015): DNA minor groove binder (treatment of specific soft tissue sarcomas: liposarcoma and leiomyosarcoma)
38	Dolutegravir (2013): Integrase inhibitor (treatment of HIV infection)
39	Bictegravir (2018): Integrase inhibitor, approved in combination with emtricitabine and tenofovir alafenamide (treatment of HIV infection)
40	Cabotegravir (2021): Integrase inhibitor, approved in combination with rilpivirine (treatment of HIV infection)

); the compounds were divided according to the complexity around the piperazine ring. Table 1 shows molecules having substituents on one (1–3) or both piperazine nitrogen atoms (4–28), and they are listed according to the kind of substitution, grouped into monoaryl (1–3), diaryl (4, 5), aryl-alkyl (6–17), dialkyl (18–24), alkyl-acyl (25, 26) and diacyl piperazines (27, 28).

Table 2 reports a smaller number of molecules, with higher complexity on the piperazine ring. In these compounds, substituents are present on one or more C-atoms. In compound 29, the piperazine ring is inserted into a 3,6-diazabicyclo[3.1.1]heptane ring. One C-position of the piperazine moiety is substituted in compounds 30–32, and in 33–40, the piperazine ring is included in a more complex polycyclic structure. An endocyclic carbonyl function characterizes compounds 38–40.

An analysis of the synthetic methods used to prepare these drugs is reported in Section 2 and Section 3, with emphasis on the chemistry involving the assembly, decoration and reactivity of piperazine or of the building block containing it. Our analysis is mainly limited to the synthetic methods applied in the medicinal chemistry and in the process chemistry routes developed by the originator company; procedures developed by other researchers or by generic industries are taken into account only in a few instances.

2. Synthesis of Drugs Carrying a Piperazine Ring Decorated Only on N-Atoms

2.1. N-Aryl Derivatives

The main synthetic methods to obtain N-arylpiperazine from aromatic compounds (often halides) and piperazine are Pd-catalyzed Buchwald–Hartwig coupling, the Cu-catalyzed Ullmann–Goldberg reaction and aromatic nucleophilic substitution (SN_Ar) on electron-deficient (hetero)arenes [9,10]. Alternatively, the piperazine ring can be built from a suitable aniline and bis-(2-haloethyl)amine or diethanolamine. These methods have been widely used in the discovery chemistry, but not all may be suitable in the synthetic procedures applied for the clinical or commercial supply, owing to possible problems in the scale-up process regarding yield, purification or safety concerns.

Palbociclib (1), Ribociclib (2), Trilaciclib (6) and Abemaciclib (21) are Cyclin-Dependent Kinase (CDK) 4/6 inhibitors showing selectivity over CDK 1,2,5,7,9. Compounds 1, 2 and 21 were approved for the treatment of metastatic breast cancer, and 6 is used to reduce myelosuppression induced by chemotherapy treatments in small cell lung cancer (SCLC). The X-ray structures of 1, 2 and 21 with CDK6 show that the compounds bind to the kinase-inactive conformation. The 2-aminopirimidine moiety interacts with the hinge region, and the positively charged piperazine ring lies in the solvent-exposed region close to Thr107 and Asp104. The interaction with the latter contributes to CDK4/6 selectivity [11]. The medicinal chemistry and synthetic approaches for the preparation of 1, 2 and 21 were recently reviewed [12].

The preparations of 1 and 2 involve the same building block, t-butyl 4-(6-aminopyridin-3-yl)piperazine-1-carboxylate (41a), which was obtained through SN_Ar starting with piperazine and 2-nitro-5-halopyridine, followed by N-protection and catalytic hydrogenation (Scheme 1) [13,14]. In the discovery chemistry, the preparation of Palbociclib involved an SN_Ar reaction of 41a with 42a (X = MeSO) in toluene, but with an unsatisfactory yield (38%) [13]. Later, the synthetic procedure was optimized; the nucleophilicity of 41a was improved by using a base (cyclohexyl magnesium chloride gave the best result) and by changing the leaving group of 42 from sulfoxide (42a) to chloride (42b, Scheme 1) [15]. The final compound 1 was obtained after Heck coupling on the bromine atom of 43 with butyl vinyl ether, followed by deprotection in an acidic medium [15,16].

Different from 1, in the first synthesis of Ribociclib (2), compound 41a was coupled with chloropyrimidine 44 through a palladium-catalyzed Buchwald–Hartwig amination reaction (Scheme 1) [17]. However, since the purification of the compound from the metal catalyst was troublesome, a transition metal-free synthesis was later developed and optimized for flow chemistry, involving the use of lithium bis(trimethylsilyl)amide (LiHMDS) as a base [18]. The synthesis of Trilaciclib (6) has evolved with similar chemistry. At first, 41b, synthesized from N-methylpiperazine and 2-nitro-5-bromopyridine, was coupled with the chloro derivative 45a using a Pd-catalyzed reaction [19], and after the optimization of the synthetic route, it was reacted with methylsulfone 45b using LiHMDS as a base [20].

Trilaciclib also contains a ketopiperazine moiety, whose synthesis is described in Scheme 2. In the first patent [19], the piperazinone ring of 45 was prepared via intramolecular condensation, forming the lactam moiety. In fact, diamine 46 (prepared in three steps from cyclohexanone) was condensed with 5-bromo-2,4-dichloropyrimidine, and on compound 47, a Sonogashira reaction with 3,3-diethoxyprop-1-yne followed by ring closure afforded the protected aldehyde 48. Elaboration of the aldehydic moiety to carboxylic ester, followed by N-Boc removal, afforded 45a. In another patent [20], 45b was prepared using a synthon, 1,4-diazaspiro[5.5]undecan-3-one 49, in which the piperazinone ring was already formed. This compound was prepared in two different ways from cyclohexanone: (1) through the condensation of 46 with methyl bromoacetate, followed by N-deprotection leading to spontaneous amide bond formation; or (2) by reacting cyclohexanone with glycine methyl ester and trimethylsilylcyanide (TMSCN) to obtain 50, which, after the nitrile reduction to amine, spontaneously formed the lactam ring, giving 49. Condensation of 49 with ethyl 4-chloro-2-(methylthio)pyrimidine-5-carboxylate gave 51; after the protection of the amidic NH with Boc anhydride, base-catalyzed intramolecular condensation afforded 52. Removal of the phenolic group through the reduction of the triflate derivative gave 53; N-deprotection and oxidation of the sulfide to sulfone finally led to 45b.

The synthesis of 21 is reported in Section 2.2.

Vortioxetine (3) is an antidepressant agent with multiple activity on the serotoninergic system. The aim of Lundbeck’s researchers was the development of an antidepressant multitarget agent able not only to block the serotonin (5-HT) transporter (SERT), but also to activate the 5-HT1A receptor to obtain rapid autoreceptor desensitization, and to antagonize the 5-HT3 receptor to exert a positive effect on mood and cognitive impairment in patients with depression [21]. Later, it was found that 3 is also a partial agonist at the 5-HT1B subtype and an antagonist at the 5-HT7 and 5-HT1D receptors [22]. Vortioxetine increases neurotransmission in brain areas associated with major depressive disorder and also displays procognitive and antihyperalgesic activity [23,24].

Vortioxetine is a relatively simple molecule that can be prepared in many different ways. Some of the synthetic procedures that have been proposed to prepare 3 are reported in Scheme 3. The original method developed at Lundbeck involved the Buchwald–Hartwig addition of N-Boc-piperazine to (2-bromophenyl)(2,4-dimethylphenyl)sulfane 54a, followed by acidic deprotection, obtaining 3 in a 66% yield [25]. Later, the same industry reported a different procedure, using commercially available 2-bromophenylpiperazine that involved the Pd-catalyzed addition of 2,4-dimethylphenylthiophenol on the bromine atom [21]. When the Buchwald–Hartwig reaction was performed between piperazine and the iodo derivative 54b, the yield increased to 95%; unprotected piperazine and 54b can also react in the presence of copper salts and a ligand (preferably 2-phenylphenol) [26]. Other methods that do not involve Pd catalysts have been described. The piperazine ring was built by reacting bis(dichloroethyl)amine with aniline 55 in a high-boiling solvent [27]; alternatively, bromo-lithium exchange on 2-bromophenylpiperazine and reaction with 1,2-bis(2,4-dimethylphenyl)disulfane afforded 3 in a 70% yield [28]. A method involving an SN_Ar reaction with piperazine is possible if an electron-withdrawing group is present on the ring (step g); after the oxidation of 54c to 56 using meta-chloroperoxybenzoic acid (m-CPBA), the reaction of the sulfoxide 56 with piperazine and subsequent reduction with magnesium in methanol afforded 3 in a 45% yield [29].

Avapritinib (4) is a potent inhibitor of mutant forms of the protein tyrosine kinase KIT and of platelet-derived growth factor receptor alpha (PDGFRA), two membrane proteins belonging to the class III receptor tyrosine kinase subfamily. Avapritinib is indicated for the treatment of adults with metastatic gastrointestinal stromal tumors (GIST) harboring the PDGFRA D842V mutation; the drug is also used to treat systemic mastocytosis characterized by a KIT variant with the D816V mutation [30]. Such mutations are located in the activation loop and cause constitutive activity in the kinases, which become resistant to inhibitors binding to the inactive conformation, such as imatinib or sunitinib [31]. Avapritinib was developed by Blueprint Medicines, starting with the screening of a library of kinase inhibitors followed by optimization. Half-maximal inhibitory concentration (IC₅₀) values of the activation loop mutants of KIT and PDGFRA reached the subnanomolar range, and the compound showed high selectivity for these proteins over several other kinases [32]. Avapritinib has also demonstrated inhibitory activity toward ABC drug transporters (ABCB1, P-glycoprotein, MDR1; ABCG2, BCRP) [33].

The synthesis of Avapritinib (Scheme 4) is reported only in patents. Both aryl groups on the piperazine N-atoms are aza-heterocycles, making SN_Ar reactions feasible. In fact, the initial intermediate 57 has been prepared in high yields by coupling N-Boc-piperazine and ethyl 2-chloropyrimidine-5-carboxylate. Transformation of the carbethoxy moiety of 57 into the Weinreb amide (58), followed by a Grignard reaction using 4-fluorophenyl magnesium bromide, gave the 4-fluorobenzoyl analog 59, which was reacted, after acidic deprotection, with the chloro derivative 60 for the second SN_Ar reaction to obtain 61. The addition of methyl magnesium bromide on the chiral sulfinamide (S)-62, followed by acidic hydrolysis and the chiral supercritical fluid chromatographic separation of the diasteroisomers, gave 4 [34]. The synthesis of 4 was later modified [35]. The procedure to insert the chiral center was performed on 59 and optimized to obtain (S)-64 in high amounts without the need for expensive chromatographic separation. The yield of the reaction between (S)-64 and 60 has been improved using ethanol in place of dichloromethane as a solvent [36].

Letermovir (5) is an antiviral agent against human cytomegalovirus (hCMV) infection. CMV is a herpes virus that causes severe morbidity and mortality in immunocompromised individuals, such as those with advanced human immunodeficiency virus (HIV) infection or those who have received solid organ or bone marrow transplants [37]. Letermovir was discovered by AiCuris through a high-throughput screening (HTS) campaign, aimed to find compounds with a mechanism of action that is different from the inhibition of DNA polymerase, which is the target of nucleoside–nucleotide analogs such as Ganciclovir or Cidofovir, commonly used to treat this infection [38]. Terminase is a fundamental protein complex in the DNA cleaving and packaging process; Letermovir is presumed to interfere with the interaction of viral concatemer DNA with the pUL56 subunit of the terminase complex [39]. Mutations at this level confer resistance to this drug, which in turn is active against viral strains resistant to CMV DNA polymerase inhibitors [37].

We included Letermovir in the diarylpiperazine group, although the quinazoline ring attached to one piperazine N-atom is partially hydrogenated. The reported synthetic procedures exploited the nucleophilic addition of commercially available 3-methoxyphenylpiperazine on a suitable acceptor, either carbodiimmide 67 (Scheme 5A, blue route) or dihydroquinazoline 73 (Scheme 5A, red route), both prepared starting with 2-bromo-6-fluoroaniline. The conditions used for the nucleophilic attack are crucial. In the discovery synthesis, carbodiimmide 67 was obtained through the Heck coupling of the starting aniline with methyl acrylate to give 65, which was converted into the imminophosphorane 66 and then reacted with 2-methoxy-5-(trifluoromethyl)phenyl isocyanate [40]. The dihydroquinazoline 69 was obtained after the prolonged heating of 67 with (3-methoxyphenyl)piperazine, which behaved as a nucleophile and as a base to perform aza-Michael cyclization without isolating the guanidine intermediate 68 (structure shown in Scheme 5B). Ester hydrolysis and enantiomer separation by means of chiral high-performance liquid chromatography (HPLC) afforded 5. Since this route was not suitable for scaling up, a new procedure was developed (Scheme 5A, red route) [41]. The starting aniline was reacted with 2-methoxy-5-(trifluoromethyl)phenyl isocyanate to give urea 70, from which the tetrahydroquinazolinone 72 was obtained through Heck coupling with methyl acrylate under basic conditions without isolating the intermediate 71. After the transformation of 72 into the chloro analog 73, the arylpiperazine group was introduced via nucleophilic addition in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and under heating. Two crystallizations of the salt of racemic 69 with (S,S) di-(p-tolyl)tartaric acid, followed by basic hydrolysis, gave Letermovir (the S enantiomer) in about a 41% yield from 72.

Later, a more efficient and cost-effective route (Scheme 5B) was developed for a long-term supply of the drug [42]. The transformation of 71 (synthesized from 65) into 67 using phosphorous pentachloride and 2-picoline as a base, with careful control of the reaction conditions for the addition of (3-methoxyphenyl)piperazine, afforded 68 in a 90% yield, avoiding the formation of racemic ring-closure by-products. Aza-Michael cyclization was performed in the presence of the cinchona alkaloid analog 74 as a chiral phase transfer catalyst at a low temperature and by using potassium phosphate as a base. (S)-69 was obtained with 76% ee, which was increased to >99% after crystallization of the (S,S) di-p-tolyltartrate salt. Basic hydrolysis afforded 5.

Other compounds whose synthesis involves an SN_Ar reaction of N-alkylpiperazine on the suitable aromatic derivative are 7–11. The aromatic substrate is a halobenzene carrying electron-withdrawing groups for 7 and 8 and an aza-heterocycle for 9–11.

Infigratinib (7) is a fibroblast growth factor receptor (FGFR) inhibitor approved in 2021 for the treatment of cholangiocarcinoma, and it is under investigation for other malignancies characterized by abnormal activity of FGFRs [43]. The design of 7 by Novartis involved a ring-opening strategy to mimic the 6-aryl-pyrido[2,3-d]pyrimidin-7-one structure already found in promising protein kinase inhibitors. The pyridone ring was replaced with a urea moiety, and the substitution pattern on the pyrimidine ring was modified to obtain a structure mimicking the original one, owing to an intramolecular H-bond (Scheme 6A) [44]. Further optimization then led to Infigratinib, endowed with selectivity for FGFR 1–3 over several other kinases [45].

Compound 7 was synthesized through a four-step procedure starting with commercially available N-ethylpiperazine and 1-bromo-4-nitrobenzene. After condensation, followed by catalytic hydrogenation, intermediate 75 was treated with 6-chloro-N-methylpyrimidin-4-amine 76. The urea linkage was formed through the reaction of 77 with the isocyanate 78, obtaining 7 in a 71% yield (Scheme 6B) [45].

Entrectinib (8) is a kinase inhibitor approved for ROS1-positive metastatic non-small cell lung cancer (NSCLC) and neurotrophic receptor tyrosine kinase gene fusion positive solid tumors. Entrectinib, with similar potency, inhibits anaplastic lymphoma kinase (ALK), ROS1 and tropomyosin receptor kinase (TRK), showing antiproliferative activity in cancers originating from gene fusion mutations involving these proteins. Entrectinib is able to cross the blood–brain barrier, being effective against primary and metastatic brain tumors [46]. Entrectinib was discovered at Nerviano Medical Sciences, starting with the HTS of a corporate compound collection. The initial hit was optimized for the interaction with ALK; a kinase selectivity screening evidenced that the compound was also active on ROS1 and TRK [47].

The synthesis of 8 (Scheme 7) started with commercially available 4-fluoro-2-nitrobenzoic acid, which was transformed into the t-butyl ester and then treated with N-methylpiperazine in excess, used as a reactant and solvent; reduction of the nitro group afforded 79. Reductive amination using tetrahydropyran-4-one, followed by a reaction with trifluoroacetic anhydride, gave 80, whose ester group was hydrolyzed under acidic conditions to 81 [48]. Treatment of 81 with oxalyl chloride gave the acyl chloride 82, which was coupled with amine 83; with this route, 8 was obtained in a 46% overall yield [47].

Avatrombopag (9) is a second-generation orally available thrombopoietin (TPO) receptor agonist approved by the FDA to treat thrombocytopenia in patients affected by chronic liver disease and scheduled to undergo an invasive procedure [49]. The design and development of 9 has not been published yet; there is evidence that 9 interacts with the TPO receptor in a binding site that is different from the endogenous agonist. In the transmembrane domain, a histidine residue at position 499 has been shown to be critical for Avatrombopag activity [50,51].

In Scheme 8, we report the synthesis of Avatrombopag as found in the original patent [52]. The synthesis started with 4-(4-chlorothiophen-2-yl)thiazol-2-amine 84, which was transformed into the 5-bromo derivative and in situ reacted with N-cyclohexylpiperazine. The coupling of 85 with 5,6-dichloronicotinic acid gave 86, which was treated with ethyl isonipecotate, obtaining 87. Basic hydrolysis afforded 9.

Netupitant (10) and Fosnetupitant (11) are neurokinin receptor 1 (NK1)-selective antagonists approved in combination with the 5-HT3 receptor blocker Palosetron to treat nausea and vomiting in patients undergoing cancer chemotherapy; the water-soluble 11, the prodrug of 10, is used as an intravenous (iv) formulation. The two-drug combination (called NEPA), administered in association with dexamethasone, proved to be effective both in the acute phase of chemotherapy-induced nausea and vomiting, mainly due to the activation of 5-HT3 receptors, and in the delayed phase, in which substance P stimulation of NK1 plays a major role [53].

Three different synthetic procedures of Netupitant are reported in Scheme 9; commercially available N-methylpiperazine is always introduced through an SN_Ar reaction on a pyridine derivative. In the discovery synthesis (Scheme 9A), this reaction was performed on 2-chloro-5-nitropyridine with quantitative yields [54,55]. The nitro group of 88 was then transformed into pivaloylamide (89) in order to direct the following ortho-metalation-iodination sequence, leading to 90. Subsequent Suzuki coupling gave 91; acidic hydrolysis of the pivaloyl protective group led to amine 92. Mono-methylation to afford 93 was achieved using a reductive ortho-ester procedure; then, the reaction with 94 gave 10 in a 15% overall yield. The drawbacks of this procedure were the lithiation reaction, for which a low temperature was necessary, the use of an expensive boronic acid reactant and a suboptimal yield.

The first process synthesis (Scheme 9B) started with commercially available 6-chloro-nicotinic acid, which was transformed into the t-butylamide 95. The key step in this route was the 1,4-addition of a Grignard reagent, followed by oxidation of the dihydropyridine intermediate (not shown in Scheme 9) to obtain pyridine 96. The addition of N-methylpiperazine was performed using the amine as a solvent; after the removal of the t-butyl group, 97 was obtained in a high yield [56]. The amide 97 was transformed into the carbamate 98 and reduced to 93; the final acylation led to 10 in a 63% overall yield.

Another procedure was also developed (Scheme 9C) [57], based on a new, fast and productive synthesis of pyridine 99; this new route can avoid the use of expensive 6-chloronicotinic acid and can bypass the Grignard addition step, whose work-up and purification are considered troublesome when applied to industrial quantities. The introduction of N-methylpiperazine into 99 was regioselective on the 6-Cl, giving 100 in a 77% yield. The 2-Cl was removed via hydrogenolysis using Pearlman’s catalyst, and the CN group was hydrolyzed to obtain amide 97, which was transformed into 10, as seen before.

Fosnetupitant 11 was prepared via alkylation of the N-methylpiperazine group with di-t-butyl (chloromethyl) phosphate 101, followed by acidic hydrolysis of the t-butyl ester moiety; this method gave 11 with a yield of about 90% (Scheme 9A) [58].

Venetoclax (12) is a mimetic of the homology domain 3 (BH3) of B Cell Lymphoma protein (BCL), which antagonizes the activity of the pro-survival protein BCL-2, leading to the apoptosis of cancer cells. Venetoclax is approved to treat patients with chronic lymphocytic leukemia with a 17p deletion who received at least one prior therapy; it is also tested in clinical trials for other hematological malignancies [59]. The discovery of a BCL antagonist was addressed by Abbott by means of fragment-based drug discovery, but the development of 12 was troublesome due to difficulties improving BCL2 selectivity and optimizing the pharmacokinetic properties of the drug candidates [60,61].

The first large-scale synthesis of Venetoclax was developed through the optimization of the reaction conditions used in the medicinal chemistry route (Scheme 10A) [62]. In the optimized procedure, unprotected piperazine was reacted with 102, prepared with commercially available methyl 2,4-difluorobenzoate and 1H-pyrrolo[2,3-b]pyridin-5-ol. To minimize the formation of the double addition product, piperazine was used in excess (8 eq). Reductive amination of aldehyde 104 with 103 gave the ester 105, which was hydrolyzed to 106 under basic conditions and coupled with sulfonamide 107, obtaining 12. Although the careful optimization of reaction conditions (time, solvent, temperature, equivalents of reactants) led to a synthetic route that was able to provide 12 on a multikilogram scale, it was not considered effective due to several drawbacks, such as a low overall yield and the formation of several impurities whose removal would increase the synthetic costs. In particular, concerns have been raised regarding the poor regioselectivity of the reaction of methyl 2,4-difluorobenzoate with hydroxyazaindole (first step in Scheme 10A). Therefore, the new route used t-butyl 2-fluoro-4-bromobenzoate 108, conveniently prepared with 4-bromo-2-fluoro-1-iodobenzene, which gave a clean reaction with 1H-pyrrolo[2,3-b]pyridin-5-ol, leading to 109 (Scheme 10B). Another concern was the instability of aldehyde 104, which complicated the purification of the intermediate compounds from impurities with mutagenic or carcinogenic potential. The main advancement in the new procedure was the involvement of a Buchwald–Hartwig amination reaction to obtain the N-arylpiperazine structure. Therefore, a freshly prepared solution of 104 was reacted with N-Boc-piperazine under a reductive amination protocol to give 110, which was hydrolyzed to 111 and coupled with ester 109 using a Pd-catalyzed reaction. In this way, 112 was obtained with a high yield and purity. After hydrolysis to 106, final coupling with 107 was performed as seen before, obtaining 12.

Brexpiprazole (13) is a dopamine 2 receptor (D2) partial agonist approved by the FDA for the treatment of schizophrenia and as an adjunctive treatment for major depressive disorder. Brexpiprazole shows high affinity for several subtypes of serotonin, dopamine and noradrenaline receptors, behaving as a partial agonist on 5HT1A and antagonist on 5HT2A [63]. Brexpiprazole is a close analog of Aripiprazole (structure 127), the first third-generation antipsychotic drug, from which it differs because of its lower intrinsic activity, which results in higher tolerability [64].

Also, the synthesis of 13 involved a Pd-catalyzed amination reaction. The first synthesis used by Otsuka Pharmaceuticals to prepare 13 looks very simple, since it consists of the separate preparation of intermediates 114 and 116 and their subsequent condensation (Scheme 11) [65]. The piperazine-containing intermediate, compound 114, was prepared starting with 113a and unsubstituted piperazine by means of a Pd-catalyzed amination reaction; 114 was obtained after chromatographic purification and was transformed into a hydrochloride salt. Compound 116 was prepared through the alkylation of 115 with 1-bromo-4-chlorobutane; the condensation of 114 with 116 using the Finkelstein reaction gave 13 in a 50% yield from 113a. Besides the low yield, steps a and b produced a relatively large number of by-products whose removal was difficult. Otsuka Pharmaceuticals improved the synthetic method, developing routes to prepare 113b starting with 2,6-dichlorobenzaldehyde [66] and trying different Pd catalysts for step a [67]. Kumar et al. reported that, although the yield of 114 was improved, the formation of by-products was not completely suppressed [68]. More recently, researchers from the Chinese Academy of Sciences developed, on a kilogram scale, a new method using 2-chloro-6-fluorobenzaldehyde. The nucleophilic displacement of fluorine with N-Boc-piperazine gave 117; on this compound, the benzothiophene ring was assembled using thioglycolic acid, followed by decarboxylation and deprotection, obtaining 114. Even if the total yield of 114 from 2-chloro-6-fluorobenzaldehyde was not high (54%), the impurities found in the formation of 117 were easily removed, yielding a high-purity compound [69].

Vilazodone (14) is a serotoninergic modulator approved for the treatment of major depressive disorder. This compound is endowed with serotonin reuptake inhibitory and 5-HT1A receptor agonist activities, with (sub)nanomolar IC₅₀ values. This compound showed high selectivity for these targets with respect to other serotoninergic subtypes and to dopaminergic, adrenergic and histaminergic receptors [70].

In the initial synthesis of Vilazodone (Scheme 12) [70], the piperazine ring was built using bis-(2-chloroethyl)amine and ethyl 5-aminobenzofuran-2-carboxylate, prepared through the reduction of the commercially available 5-nitro derivative. Compound 118 was then reacted with 3-(4-chlorobutyl)-1H-indole-5-carbonitrile 119a, prepared with commercially available 5-cyanoindole; finally, the carbethoxy group of 120a was transformed into carboxamide to obtain 14. This procedure had several drawbacks due to the formation of various by-products and the use of expensive reagents, leading to a low yield, difficult purification and high costs. The procedure was then changed by protecting the indole NH with tosyl amide and by optimizing the reaction conditions due to the presence of the protecting group. In the reaction between 118 and 119b, potassium iodide was added to improve the yield. The basic hydrolysis of the carbethoxy moiety, necessary for its transformation into carboxyamide, also removed the tosyl group [71].

Flibanserin (15) is a 5-HT1A receptor agonist approved for the treatment of low sexual desire disorder in premenopausal women. In vitro, this compound binds with high affinity to the 5-HT1A, dopamine D4 and 5-HT2A receptors, on which Flibanserin behaves, respectively, as an agonist, a very weak partial agonist and an antagonist [72]. A decrease in the serotonergic inhibition of excitatory neurotransmitters, dopamine and norepinephrine, is supposed to be the basis of its activity in female sexual desire. Flibanserin was originally developed as a treatment for depression [73].

The original synthesis of 15 started with commercially available 1-[3-(trifluoromethyl)phenyl]piperazine, which was alkylated with 1-(2-chloroethyl)-1,3-dihydro-2H-benzo[d]imidazol-2-one 121 to give 15 (Scheme 13, yield not indicated in the patent) [74]. Subsequent efforts mainly regarded the optimization of the synthesis of 121 (see, for instance, ref. [75]). More recently, 15 was synthesized also by means of a new general method for the preparation of 1,4-disubstituted piperazines that exploits quaternary N-aryl-1,4-diazabyciclo[2.2.2]octane salts [76]. The addition of 1,4-diazabyciclo[2.2.2]octane (DABCO) to diaryliodonium triflate 122 gave the ammonium derivative 123, which was reacted with benzimidazolone 124 (sodium salt) to obtain 15 after the acidic removal of the isopropenyl group.

Aripiprazole Lauroxyl (16) and Cariprazine (17) are antipsychotic drugs used in the treatment of schizophrenia. Compound 16 is the prodrug of Aripiprazole, a drug used since 2002 for the treatment of a wide variety of mood and psychotic disorders, and it has been developed into a long-acting injectable form. Compound 16 is metabolized into N-hydroxymethyl Aripiprazole and lauric acid; the former is then hydrolyzed into formaldehyde and Aripiprazole, achieving the maximal concentrations of the drug after about 41 days. This prodrug demonstrated improved medication adherence and reduced relapse rates [77,78]. Cariprazine 17 was obtained from the optimization of an impurity isolated during the large-scale synthesis of RG-15, an antipsychotic under development by Gedeon Richter and Forest Laboratories [79]. This compound is a D2/D3 receptor antagonist–partial agonist with preferential affinity for the D3 subtype; moreover, it also displays high affinity for the serotoninergic 5-HT2B receptor [80].

Both 16 and 17 were prepared starting with the same building block 1-(2,3-dichlorophenyl)piperazine 125. In the original synthesis of Aripiprazole (Scheme 14) [81], 125 was prepared starting with 2,3-dichloroaniline and diethanolamine (according to ref [82], details are not given); reaction with the bromoalkyl derivative 126 gave Aripiprazole (127), which was converted into 16 via treatment first with formaldehyde and then with a lauric acid derivative (not indicated in the patent, presumably lauroyl chloride) [83].

For the synthesis of Cariprazine, Gedeon Richter and Forest Laboratories prepared 125 by means of a Buchwald–Hartwig reaction of N-Boc-piperazine on 1-bromo-2,3-dichlorobenzene, followed by acidic hydrolysis (Scheme 14) [84,85]. The N-Alkylaton of 125 was accomplished via reductive amination using aldehyde 128a or via the alkylation of the mesylate 128b, yielding 129. After deprotection, the urea moiety was formed via reaction with dimethylcarbamic chloride [86,87]. Alternatively, amine 125 was acylated with acid 130 and 1,1′-carbonyldiimidazole (CDI), and the amide group of 131 was selectively reduced using sodium borohydride and boron trifluoride etherate to furnish 17 [88]. The routes that use aldehyde 128a or mesylate 128b have similar yields (about 35%), and that involving acid 130 seems more efficient (57% yield). A recent paper reviewed all the synthetic methods used to prepare 17, which mainly differ in their preparation of the cyclohexyl building blocks (128 or 130) [85].

2.2. N-Alkyl Derivatives

There are three important methods to transform piperazine into its N-alkyl analogs: nucleophilic substitution on alkyl halides or sulfonates, reductive amination [89] and the reduction of carboxyamides [90]. N-alkylpiperazines 12–23 have been prepared using these procedures. As we have just seen, all three of these methods have been used for the preparation of 17 using different synthetic routes developed by the originator company (Gedeon Richter and Forest Laboratories) (Scheme 14).

N-alkylation using alkyl chlorides or bromides has been used for the synthesis of compounds 13–16 and 18, shown in Scheme 11, Scheme 12, Scheme 13, Scheme 14 and Scheme 15. In these cases, the addition of sodium or potassium iodide, which promotes halogen exchange and increases the reactivity of the leaving group, was performed in order to improve the yield (see, for instance, the synthesis of 14, Scheme 12) or to avoid reaction conditions that are too harsh.

N,N′-dialkylpiperazine 18–23 are kinase inhibitors used in the treatment of various types of cancers. Bosutinib (18) is a dual ABL/SRC inhibitor, and Ponatinib (19) is a BCR-ABL inhibitor. Both drugs are used for Imatinib-resistant chronic myelogenous leukemia (CML); different from 18, 19 is able to inhibit the T315I mutant enzyme [91]. Nintedanib (20) was approved for the treatment of NSCLC and for idiopathic pulmonary fibrosis. Nintedanib inhibits different angiokinases (PDGFR, FGFR and vascular endothelial growth factor (VEGFR)) but also non-receptor kinases [92].

In the synthetic routes originally developed at Wyeth for Bosutinib (18), the piperazine ring was inserted in the last synthetic steps, in two different ways (Scheme 15): through the nucleophilic attack of N-methylpiperazine on the 3-chloropropyl derivative 132 using the amine as a solvent, or through the addition of commercially available 1-(3-hydroxypropyl)-4-methylpiperazine on the aryl fluoride 133 using sodium hydride to increase the nucleophilicity of the OH group [93,94]. The procedure used for the manufacture of late-stage clinical supplies, on the contrary, inserted the piperazine moiety as the first step via the reaction of N-methylpiperazine with the chloroalkyl derivative 134, obtaining 135, which was reduced to 136. The quinoline ring of 18 was built on the amino group via reaction with cyanoacetamide 137 and triethyl orthoformate, obtaining 138 as a mixture of cis/trans isomers; cyclization using phosphorus oxychloride in acetonitrile provided 18 [95]. In both routes, the reaction of N-methylpiperazine was facilitated by the addition of sodium iodide.

The syntheses of Ponatinib (19) and Nintedanib (20) required piperazine N-alkylation using reactive alkyl halides; therefore, there was no need to add an iodide salt. The original procedures involved the addition of commercially available N-methylpiperazine on benzyl bromide 139 or on bromoacetyl derivative 143 (Scheme 16). For the synthesis of 19, benzyl bromide 139 was prepared via the bromination of commercially available 1-methyl-4-nitro-2-(trifluoromethyl)benzene followed by the reaction with N-methyl-piperazine, obtaining 140. Reduction of the nitro group and the reaction of amine 141 with 3-iodo-4-methylbenzoyl chloride gave 142, which was coupled with 3-ethynylimidazo[1,2-b]pyridazine under Sonogashira conditions, obtaining 19 [96]. Compound 143 was prepared through the acylation of N-methyl-4-nitroaniline with bromoacetyl bromide. After the addition of N-methylpiperazine, the nitro group of 144 was hydrogenated, and aniline 145 was coupled with 146 to yield 20 [97,98].

Also, the synthesis of Maralixibat (24) involves N-alkylation using a reactive alkyl halide; in this compound, the piperazine ring is part of a DABCO moiety, with one N-atom being quaternary. Maralixibat is an ileal bile acid transporter (IBAT) inhibitor approved for the treatment of rare cholestatic liver diseases, including Alagille syndrome, progressive familial intrahepatic cholestasis and biliary atresia. Maralixibat is a quaternary ammonium compound that is minimally absorbed and interacts with the transporter in the ileal lumen. IBAT inhibition reduces bile acid reabsorption and increases their elimination with feces. As a consequence, the serum levels of bile acids are reduced, and so is the risk of bile acid-mediated liver damage [99].

The most laborious part of the synthesis of 24 is related to the preparation of the chiral benzothiazepine oxide (4R,5R)-147 (not addressed), and the insertion of the piperazine moiety simply involved the nucleophilic addition of DABCO to the reactive benzylic chloride 149 (Scheme 17), prepared via the alkylation of compound 147 with 1,4-α,α′-dichloro-p-xylene 148. The quaternary ammonium nature of 24 facilitated its purification since this compound readily precipitated from the reaction mixture [100,101].

Reductive amination has been used to prepare the N-alkyl compounds 12, 16, 21 and 22 using the suitable aldehyde and sodium triacethoxyborohydride. The syntheses of 12 and 16 are reported in Scheme 10 and Scheme 14, respectively, and those of 21 and 22 are shown in Scheme 18 and Scheme 19, respectively.

The CDK 4/6 inhibitor Abemaciclib (21) is described in Section 2.1. Its synthesis (Scheme 18) started with commercially available 6-bromonicotinaldehyde. Reductive amination using sodium triacethoxyborohydride and N-ethylpiperazine gave 150, which was transformed into aniline 151 and reacted with 2-chloropyrimidine 152, yielding 21 [102]. Alternatively, compound 152 was condensed with 6-aminonicotinaldehyde using Pd-catalyzed coupling, and then 153 was treated with N-ethylpiperazine using Leuckart–Wallach conditions [103,104]. In this case, the use of triacethoxyborohydride was discarded since it produced a small percentage of a by-product, the alcohol deriving from the reduction of aldehyde 153, which was difficult to eliminate during purification.

Gilteritinib (22) is indicated for the treatment of acute myeloid leukemia deriving from an FMS-like tyrosine kinase gene (FLT3) mutation detected by a companion diagnostic; it is a dual FLT3/AXL selective inhibitor [105]. Brigatinib (23) is an ALK inhibitor, carrying an unusual dimethylphosphine oxide moiety as an H-bond acceptor [106]; it is approved to treat ALK-positive metastatic NSCLC characterized by the presence of an EMAP Like 4 (EML4)−ALK fusion protein. Both Gilteritinib and Brigatinib carry the 1-methyl-4-(piperidin-4-yl)piperazine group; their original synthetic routes used this commercially available reagent (Scheme 19 and Scheme 20).

The synthesis of 22 was reported by Astellas only in patents, from which it is difficult to extract the relevant information on reaction conditions and yields [107]. The description of the original route was obtained from another source [108], which reports that the reaction of compound 154 with pyrazine 155 at high temperatures (150–200 °C), followed by the treatment of 156a with tetrahydro-2H-pyran-4-amine, gave 22 in only a 25% yield (Scheme 19). Differently, the method reported in patent CN106083821A [108] involved the Pd-catalyzed amination of 157 with 154, obtaining 156b; the hydrolysis of the nitrile group to carboxyamide gave 22. The total yield was higher with the new route. The preparation of 154 (reaction of 1-methyl-4-(piperidin-4-yl)piperazine with 1-fluoro-2-methoxy-4-nitrobenzene followed by catalytic hydrogenation) was reported by Astellas in ref. [109]

For the synthesis of 23 (Scheme 20), similar chemistry was applied to obtain 158, which was reacted with 2-chloropyrimidine 159 to obtain the final compound [106]. Very recently, the ¹¹C analog of 23 was described as a positron emission tomography (PET) radiotracer to assess the mutational status of Brigatinib’s target kinases and to predict the benefit from the treatment of NSCLC patients [110]. A free NH group on piperazine was required before making the final radiolabeling step; therefore, the 4-piperidinyl-piperazine moiety was built by reacting 4-fluoro-2-methoxy-1-nitrobenzene with 4-piperidone, obtaining 160, and by performing reductive amination using N-Boc-piperazine and sodium cyanoborohydride in the presence of acetic acid. The nitro group of 161 was hydrogenated, and 162 was coupled with 159 in the usual way. The alkylation of 163 with (¹¹C)-methyl iodide gave the desired compound (10% radiochemical yield).

The preparation of the last two N-alkylpiperazine derivatives (Mitapivat 25 and Zavegepant 26) is described in Section 2.3.

2.3. N-Acyl Derivatives

Compounds 25–28 were prepared by reacting the appropriate piperazine derivative with a suitably activated carboxylic acid. These procedures were straightforward but required optimization specially for large-scale processes.

Mitapivat (25) is a pyruvate kinase activator approved to treat hemolytic anemia in pyruvate kinase (PK) deficiency. PK catalyzes the final step in glycolysis, converting phosphoenolpyruvate to pyruvate, producing adenosine triphosphate (ATP). Mitapivat was obtained starting with quinoline-8-sulfonamides I (Scheme 21A), an activator of the PK-M2 isoform found in muscle cells [111]. The pyrazine moiety was replaced with a cyclopropylmethyl group in order to obtain activating properties on the isoform expressed in the red blood cell (PK-R). Mitapivat binds to an allosteric site, distinct from the pocket occupied by fructose bisphosphate (the natural enzyme activator), and it is able to trigger both the wild type and the mutated isoforms [112,113].

The synthesis of Mitapivat is shown in Scheme 21B. Quinoline-8-sulfonyl chloride was reacted with ethyl 4-aminobenzoate, and the ester function was hydrolyzed to give 164. Amide 165 was prepared by coupling 164 with N-Boc-piperazine using (benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate (PyBoP) as a coupling reagent, followed by acidic deprotection. The reaction of 165 with cyclopropanecarbaldehyde and sodium triacetoxyborohydride gave the final compound with a yield below 45% [114]. In a following patent, Agios Pharmaceuticals improved the process by reacting 164 with CDI and 1-(cyclopropylmethyl)piperazine 166 (prepared by means of reductive amination with N-Boc-piperazine and cyclopropanecarbaldehyde) [115].

Zavegepant (26) is a calcitonin gene-related peptide (CGRP) receptor antagonist with high affinity for its receptor (Ki 23 pM) and high potency in reverting the CGRP-induced dilation of ex vivo human intracranial arteries (EC₅₀ 880 pM). Zavegepant is administered as a nasal spray formulation, owing to its good water solubility, and it is approved for the acute and/or preventive treatment of migraines [116].

The final steps of the optimized synthesis of 26 are shown in Scheme 22. Considerable efforts have been spent on the synthesis of the chiral aminoacid R-167 [117] and of the piperidine 168 [118], which were coupled using CDI to afford urea 169 after the hydrolysis of the carbomethoxy group. The insertion of the piperazine moiety was the last step, using commercially available 1-(1-methylpiperidin-4-yl)piperazine, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and 1-hydroxybenzotriazole (HOBt) as coupling reagents. Whereas the synthesis of urea R-169 required careful optimization, due to the ambident nucleophilicity of 167, the last step was more straightforward. To maximize process throughput, the final product (as HCl salt) was precipitated directly from the reaction mixture in N,N-dimethylacetamide (DMAc) by adding acetone [117].

Olaparib and Fostemsavir are diacylpiperazine derivatives. Olaparib (27) is a poly-ADP-ribose-polymerase (PARP) inhibitor approved for the treatment of Breast-Related Cancer Antigens (BRCA)-associated tumors, such as ovarian, breast, pancreatic and prostate cancer. PARP inhibition induces synthetic lethal interactions in cancer cells already deficient in DNA repair mechanisms, resulting in cell death. Olaparib interacts with the nicotinamide-binding pocket of the enzyme; the phtalazinone ring mimics the amide group of nicotinamide [119]. The main role of the cyclopropyl(piperazin-1-yl)methanone moiety is to increase water solubility and oral bioavailability [120], but the carbonyl group is also engaged in a water-mediated H-bond with the target.

For the synthesis of Olaparib, acid 172 was prepared starting with 3-dimethoxyphosphoryl-3H-2-benzofuran-1-one 170 through condensation with 2-fluoro-5-formylbenzonitrile, followed by the hydrolysis of the nitrile group of 171 and treatment with hydrazine hydrate (Scheme 23). The piperazine moiety was attached by reacting 172 with cyclopropyl(piperazin-1-yl)methanone in acetonitrile or by using N-Boc-piperazine with the same coupling reagents but in a different solvent (DMAc), followed by the deprotection of 173 and the reaction of 174 with cyclopropanecarbonyl chloride [121,122]. This second route was also applied to the synthesis of Olaparib derivatives carrying groups that were different from the cyclopropanecarbonyl one, which could be potentially useful as imaging agents [123].

Fostemsavir (28) is the prodrug of Temsavir (177a, structure shown in Scheme 24), an HIV attachment inhibitor, which blocks the interaction of the surface envelope protein gp120 with the CD4 receptor, thus preventing virus entry into the cells; this mechanism of action is different from that of other virus entry inhibitors such as Enfuvurtide and Maraviroc [124]. Fostemsavir is endowed with high aqueous solubility; its administration increases the plasma levels of Temsavir with respect to the administration of the parent molecule. The prodrug is hydrolyzed by an alkaline phosphatase in the gut, releasing the parent drug before absorption, resulting in very low systemic exposure of 28 [125].

The synthesis of Fostemsavir is shown in Scheme 24. Several routes are described, which require careful optimization for scaling up, owing to the complex reactivity of the azaindole moiety (see ref. [125] and the references cited therein). Here, only the procedure used in the medicinal chemistry and the commercial one are discussed. In the discovery chemistry [126], 28 was obtained starting with 2-aminopicoline, which was transformed into 175; Friedel–Crafts acylation was performed on this compound using methyl 2-chloro-2-oxoacetate. The hydrolysis of ester 176 and reaction with commercially available N-benzoylpiperazine using EDC as a coupling reagent gave amide 177a (Temsavir). Alkylation with di-t-butyl (chloromethyl) phosphate 101 and the hydrolysis of the ester groups of 178 gave 28, which was precipitated as tromethamine salt via the addition of acetone.

Differently, the optimized industrial synthesis started with 1-(phenylsulfonyl)-pyrrole, which was converted into the bromo-derivative 179, suitable for the acylation reaction necessary to insert the oxalyl moiety. The safety concerns connected to the large-scale use of nitromethane (as in step a, Scheme 24) were solved by using tetra-n-butylammonium hydrogen sulfate to increase solvent polarity and favor AlCl₃ dissolution. After the hydrolysis of the ester group, amidation was performed using diphenylphosphinic chloride (DPPCl) as a coupling reagent, obtaining 180. Regioselective copper-mediated Ullmann–Goldberg–Buchwald coupling introduced the triazole group on 180, and the addition of lithium iodide allowed obtaining 177b, which, different from 177a, was crystalline and easier to purify. Tetraethylammonium iodide was added in the reaction with 101, owing to the low reactivity of 177b in this step. In the majority of the synthetic routes developed at Bristol Myers Squibb, the attachment of N-benzoylpiperazine was performed starting with the acid and using a suitable coupling reagent. In one route, 177a was prepared by reacting ester 176 directly with the sodium salt of N-benzoylpiperazine to increase nucleophilicity (Scheme 24) [127]. Even though the yield reported for this step was high, the method involving ester hydrolysis was preferred in the optimized route.

3. Synthesis of Drugs Carrying C-Substituents on the Piperazine Ring

Commercially available piperazine reagents with C-substituents have been used to prepare compounds 29–31; therefore, the synthetic routes involve the same methods discussed previously for N-alkyl and N-aryl derivatives.

Selpercatinib (29) is a rearranged during transfection (RET) receptor tyrosine kinase inhibitor approved for cancers harboring RET mutations (metastatic RET fusion-positive NSCLC, advanced/metastatic RET-altered medullary thyroid cancer and papillary thyroid carcinoma). Selpercatinib binds to the ATP site; its unique mode of binding allows the interaction with the RET protein to avoid steric clashes with gatekeeper mutations at V804. However, mutations in other parts of the protein confer resistance to this drug [128,129].

In 29, the piperazine ring is included in a diazabicycloheptane group; the two different synthetic routes developed by Array BioPharma exploited an SN_Ar reaction to insert this moiety. Suzuki coupling between pyrazolo[1,5-a]pyridin-4-yl trifluoromethanesulfonate 181 and 2-fluoro-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine 182 gave 183, which was reacted with commercially available N-Boc-3,6-diazabicyclo[3.1.1]heptane, obtaining 184 (Scheme 25A). Removal of the protecting group and reductive amination using 6-methoxynicotinaldehyde gave 29. Alternatively, 184 was prepared starting with the dibromo derivative 186 (Scheme 25B). Borolane 185, obtained by condensing N-Boc-3,6-diazabicyclo[3.1.1]heptane with fluoropyridine 182, was reacted with 186, and another Suzuki reaction was performed to replace the 6-bromine of 187 with OH. The alkylation of phenol 188 with 2,2-dimethyloxirane gave 184. The two methods afforded 29 with a similar yield (about 41% starting with 181 or 186) [130,131].

Risdiplam (30) is an orally active splicing modifier, approved for the treatment of spinal muscular atrophy. It promotes the inclusion of the survival of motor neuron 2 (SMN2) exon 7 in the messenger RNA to produce a functional SMN protein in individuals lacking the SMN1 isoform. In 30, the 4,7-diazaspiro[2.5]octane moiety has the dual role of optimizing physicochemical properties (basicity, lipophilicity) and contributing to binding with the target molecules through the interaction of the protonated NH atom [132,133].

In the medicinal chemistry synthesis of 30, the piperazine moiety was attached as the last step by reacting commercially available unprotected 4,7-diazaspiro[2.5]octane with 189 (Scheme 26). This reaction occurred with a low yield, and so did the entire synthetic pathway (about 5% overall) [132]. Later, two optimized routes were patented [134,135]. In the first one, t-butyl 4,7-diazaspiro[2.5]octane-4-carboxylate was coupled with 2-nitro-5-bromopyridine, yielding 190; the nitro group was hydrogenated using platinum on charcoal as a catalyst with a percentage of vanadium to avoid the formation of partially reduced intermediates [136]. Compound 191 was condensed with di-t-butyl malonate to build the pyrido[1,2-a]pyrimidin-4-one ring, and the 2-hydroxy group was esterified with tosyl chloride. Suzuki coupling with 194, followed by the removal of the Boc-protecting group, gave Risdiplam with a yield of about 40%. In the final route, the yield was further improved. Compound 191 was reacted with Meldrum’s acid derivative 195; the addition product 196 was boiled in n-propanol, resulting in the formation of the pyrimidine ring. Acid 197 was heated in n-propanol containing HCl to perform decarboxylation and deprotection, obtaining 30 in about a 60% yield.

Sotorasib (31) and Adagrasib (32) are KRAS^G12C inhibitors approved for the treatment of NSCLC. Kirsten rat sarcoma viral oncogene homolog (KRAS) is an oncogene frequently mutated in various types of cancer; the protein was considered undruggable until the crystal structure of the G12C mutant was solved, revealing a previously unknown pocket suitable for small molecule binding [137]. Sotorasib and Adagrasib carry an acryloyl residue that reacts with the nucleophilic C12 side chain, forming a covalent bond.

In both compounds, the piperazine ring is decorated with a C-substituent, a methyl group for 31 and a cyanomethyl group on 32. The commercial route used to prepare 31 (Scheme 27) [138] is an optimization of the first published method [139], with the overall yield of the five-step process improved from 38 to 65%. The synthesis of the starting material 198 in its enantiomeric form (M-198) was reported in ref. [140]. The treatment of quinazolindione 198 with phosphorous oxychloride gave the chloro derivative 199; without being isolated, it was reacted with commercially available t-butyl (S)-3-methylpiperazine-1-carboxylate, obtaining 200. Pd-catalyzed Suzuki−Miyaura cross-coupling using boroxine 201 led to the piperazine derivative 202. Deprotection, a reaction with acryloyl chloride, and recrystallization from ethanol-water gave 31.

The synthesis of Adagrasib involves the use of (S)-2-(piperazin-2-yl)acetonitrile, which is not commercially available. Mirati Therapeutics described different syntheses (Scheme 28) for this building block as a N¹-Cbz derivative ((S)-205) or as dihydrochloride salt ((S)-209) [141,142,143]. The preparation of compound (S)-205 started with the protection of the NH group of t-butyl (R)-3-(hydroxymethyl)piperazine-1-carboxylate as a Cbz-derivative. The reaction of (R)-203 with mesyl chloride, followed by treatment with NaCN, gave the protected piperazine (S)-204. Removal of the Boc protective group afforded (S)-205 in a 44% yield. Compound (S)-209 was prepared using the same starting molecule. A reaction with thionyl chloride afforded the oxathiazole oxide (R)-206, which, after oxidation to (R)-207, was reacted with potassium cyanide. Removal of the protective group of (S)-208 under acidic conditions gave the desired piperazine (S)-209 as dihydrochloride salt in a 23% yield. For the commercial supply, a different procedure was applied. The reaction of N,N¹-dibenzylethanediamine with (S)-epichlorohydrin afforded (R)-(1,4-dibenzylpiperazin-2-yl)methanol 210a, containing some 1,4-dibenzyl-1,4-diazepan-6-ol 210b. The mixture was not separated; it was transformed into chloride 211a,b by a reaction with mesylchloride and lithium chloride and then into cyanide 212a,b by a reaction with sodium cyanide in DMF. The desired (S)-212a was purified from the mixture as (S)-mandelate salt, which was converted into the free base under basic conditions. The protective benzyl groups were finally removed using 1-chloroethyl chloroformate and N,N-diisopropylethylamine (DIPEA), affording (S)-209 in an 88% yield.

The synthesis of Adagrasib continued as shown in Scheme 29. The first published synthesis [142] started with tetrahydropyrido[3,4-d]pyrimidine 213a (Scheme 29A, upper part), which was transformed into (S)-214. The piperazine moiety was inserted in the penultimate step by reacting compound (S)-209 with triflate (S)-215a, in turn obtained by treating phenol (S)-214 with triflic anhydride. The amidation of (SS)-216 was then performed using 2-fluoroacrylic acid in the presence of a base (triethylamine) and of propanephosphonic anhydride (T3P) as an activating agent. This synthetic route yielded 32 with an overall yield of about 1%, starting with 213a. To support the clinical studies and the initial commercial supply, another route was developed (Scheme 29A, bottom) [143], in which the piperazine group was inserted earlier. The reaction of the Boc-protected 213b with (S)-205 (as fumarate salt) led to (S)-217 with a high yield and complete regioselectivity. The prolinol moiety was inserted by means of Pd-catalyzed C−O bond formation, and the Boc deprotection of (S,S)-218 was achieved under acidic conditions, isolating (S,S)-219 as L-tartrate salt. The naphthyl group was inserted by means of the Buchwald−Hartwig amination reaction; (S,S)-220 was purified as tosylate salt. Removal of the Cbz group was accomplished using 2-mercapto-1-ethanol, and the amidation of (S,S)-216 was performed using the previously employed activating agent (T3P) and sodium 2-fluoroacrylate. The sodium salt was used to minimize the decomposition of the acid, avoiding the addition of a base and improving the yield of this last step to 89% (70% after crystallization). The new process gave Adagrasib with an overall yield of 32%.

However, some criticisms for this route are related to the total cost due to the use of expensive Pd reagents, the early introduction of the costly chiral piperazine and the need to apply several protection/deprotection steps; therefore, a new process was developed (Scheme 29B) [144]. Intermediate (S)-214 was prepared starting with 8-chloronaphthalen-1-amine to build first the piperidine ring and then the pyrimidine one, and then the prolinol appendage was attached via an SN_Ar reaction. The chiral piperazine was inserted through another SN_Ar reaction. The treatment of (S)-209 with the 2-nitrobenzenesulfonyl ester (S)-215b and DIPEA avoided the formation of by-products deriving from nucleophilic attacks at the sulfur atom of the sulfonate group. The amidation step was performed as before without the need for final crystallization, owing to the high quality of the obtained material. This synthetic route did not involve the use of transition metals or protective groups. The five linear steps gave 32 in a 45% yield.

Fezolinetant (33) is an NK3 antagonist approved for the treatment of moderate to severe hot flashes caused by menopause. By inhibiting the binding of neurokinin B to NK3 receptors, Fezolinetant blocks the hypothalamic pituitary gonadal (HPG) axis, being effective in the treatment of sex hormone disorders [145]. In 33, the piperazine ring is inserted into a [1,2,4]triazolo[4,3-a]pyrazine moiety; the stereogenic “magic methyl” group in position 8 is important, as it influences the amide orientation in the bioactive conformation, thus improving the activity with respect to the unsubstituted analog [146].

The synthesis of 33 started with commercially available (R)-4-(2,4-dimethoxybenzyl)-3-methylpiperazin-2-one (Scheme 30), which was transformed into the piperazinoimidate 221 using Meerwein conditions (triethyloxonium tetrafluoroborate and sodium carbonate), and it treated with 3-methyl-1,2,4-thiadiazole-5-carbohydrazide 222 to build the triazole ring. The deprotection of 223 under acidic conditions and the acylation of the secondary amine 224 furnished 33. The dimethoxybenzyl moiety was employed since racemization problems were encountered when using a different protective group (i.e., Boc). Moreover, the presence of the 2,4-dimethoxybenzyl (DMB) group shortened the reaction time in the condensation reaction [146,147]. Later, a patent reported that the insertion of the benzoyl moiety as an initial step could also afford the desired compound with minimal racemization. Thus, commercially available (R)-3-methylpiperazinone was treated with 4-fluorobenzoyl chloride; amide 225 was transformed into imidate 226, which was treated with hydrazide 222 using 4-methylmorpholine as a base, obtaining 33. This second route was faster and avoided the use of protective groups; however, the two methods gave similar yields (about 40%) [148].

Lumateperone (34) is an atypical antipsychotic used to manage both positive and negative symptoms in patients with schizophrenia. Lumateperone is able to directly modulate serotoninergic and dopaminergic transmission and indirectly modulates the glutamatergic system. It shows subnanomolar affinity for 5-HT2A receptors and interacts with nanomolar affinity with the D2 receptor and the serotonin transporter. Cellular assays revealed antagonistic properties on both receptors and SERT [149].

In 34, the piperazine ring is inserted into a tetracyclic pyrido[4’,3’:4,5]pyrrolo[1,2,3-de]quinoxaline. Three different syntheses have been published for this compound [150,151]. In the first one (Scheme 31A), the starting material (3,4-dihydroquinoxalin-2(1H)-one) already contained a piperazinone ring. Treatment of this compound with sodium nitrite gave the nitroso derivative 227. Reduction using zinc in acetic acid and the condensation of the resulting hydrazine derivative with ethyl 4-oxopiperidine-1-carboxylate gave the tetracyclic carbamate 228, which was reduced on the indole double bond using sodium cyanoborohydride, obtaining rac-229. Alkylation of the secondary amide nitrogen atom, followed by the selective reduction of the amide moiety of rac-230, gave amine rac-231. After the hydrolysis of the carbamate group, the alkylation of 232 with 4-chloro-4′-fluoro-butyrophenone gave a racemic compound whose enantiomers were separated by chiral HPLC, obtaining 34 (4aS,9bR isomer). The second route (Scheme 31B) provided a faster way to obtain racemic 230. The indole ring was first built by condensing 2-bromophenyl hydrazine with 4-piperidone, and the indole double bond was reduced using triethylsilane and trifluoroacetic acid, obtaining the racemic cis indoline 233. After the protection of the piperidine amine as carbamate, a Buchwald−Hartwig cross-coupling reaction of rac-234 with benzophenone immine afforded rac-235. The alkylation of the indoline NH with ethyl bromoacetate, followed by the acidic hydrolysis of the immine, led to the formation of the piperazinone ring. After the methylation of the amide moiety, rac-230 was obtained and transformed into the final compound as before.

Later, a patent reported another route (Scheme 31C) in which the enantiomeric separation was performed in an earlier step on compound 233 by means of fractional crystallization of the mandelate salts or by means of chiral preparative HPLC [151]. In both cases, a high enantiomeric excess was obtained (yields were not indicated). After the usual protection of the piperidine NH group as ethyl carbamate (4aS,9bR 234), the indoline NH moiety was alkylated using N-methyl-2-chloroacetamide, and on intermediate 236, the closure of the piperazinone ring was achieved using a Cu-catalyzed Ullmann–Goldberg reaction. The reduction of 4a(S),9b(R)-230 and the removal of carbamate was performed as before, and the final alkylation was performed using DIPEA in 3-pentanone, with a substantial improvement in the yield.

Fosdenopterin (35) is a molybdenum cofactor precursor used to reduce the risk of mortality in patients with molybdenum cofactor deficiency type A. Mutations of the molybdenum cofactor synthesis 1 gene reduce the endogenous availability of the cyclic pyranopterin monophosphate 35, causing reduced molybdenum cofactor production. Fosdenopterin is administered as a supplement to overcome the cofactor deficiency [152].

In 35, the piperazine ring is inserted into a pyranopterin structure, with the two nitrogen atoms not carrying substituents. Such a heterocyclic structure was built starting with 2,5,6-triamino-3,4-dihydropyrimidin-4-one 237 and D-galactose phenylhydrazone 238 by means of a Viscontini reaction (Scheme 32) [153]. The treatment of 239 with an excess of Boc anhydride resulted in the formation of a mixture of hepta- and hexa-Boc-protected derivatives whose carbonate groups were hydrolyzed via the addition of sodium methoxide to achieve selective cleavage with respect to carbamates. Cyclic phosphate ester 241 was obtained after the treatment of 240 with methyl dichlorophosphate. Swern oxidation of the C-6 secondary alcohol gave the corresponding ketone, which was not isolated due to instability but was treated with trimethylbromosilane (TMSBr), achieving the removal of all protective groups (carbamates and ester) and furnishing 35.

Trabectedin (37) is a natural compound derived from the Caribbean Sea squirt Ecteinascidia Turbinata. Lurbinectedin (36) is a synthetic derivative. Both compounds are DNA minor groove binders approved as anticancer drugs for different malignancies (soft tissue sarcomas and ovarian cancer for 37 and metastatic SCLC for 36) and are presently used in clinical trials for other kinds of tumors [154]. Both compounds irreversibly bind DNA through the reaction of the aminoemiacetal functionality with the exocyclic guanine NH₂ group in guanine-cytosine-rich sequences, as demonstrated for 37 [155].

Since the natural availability of 37 is limited, these molecules are produced via chemical synthesis. The compounds contain two tetrahydroisoquinoline moieties whose N-atoms are also part of a piperazine ring. Several total syntheses of 37 have been reported in the literature (see [156] and references cited therein), in which the pentacyclic core is assembled first, followed by the 10-membered lactone moiety. Since a discussion of the various routes is out of the scope of this review, we took into consideration only two of them, limiting the analysis to the building of the piperazine ring. The industrial method to produce Trabectedin and Lurbinectedin starts with Cyanosafracin B (Scheme 33A), already containing the pentacyclic core of these compounds [157]. In the historical route developed by Corey in 1996, the piperazine ring was assembled through an internal Mannich reaction between the carbamate N-atom and the lactol group of 242 after the removal of the t-butyldimethylsilyl groups (Scheme 33B). In this way, the piperazine and the second tetrahydroisoquinoline rings were formed at the same time, obtaining 243 [158]. In 2019, researchers from the Chinese Academy of Sciences developed a new route on a multigram scale, claimed to be efficient and scalable [156]. In this method, the piperazine ring was assembled starting with 244. Swern oxidation of the hydroxymethyl group to aldehyde, N-Boc removal and an intramolecular Strecker reaction gave piperazine 245 in a good yield (Scheme 33B). Both 243 and 245 were then converted into lactone 246 and then transformed into the final compounds 36 and 37.

Dolutegravir (38), Bictegravir (39) and Cabotegravir (40) are integrase inhibitors approved for the treatment of HIV infection. These compounds share the common N-benzyl-9-hydroxy-1,8-dioxo-1,3,4,8-tetrahydro-2H-pyrido[1,2-a]pyrazine-7-carboxamide structure, and they differ in the number of fluorine atoms on the benzyl carboxyamide group in 7 and in the size and absolute configuration of the oxygenated heterocycle fused to the N2-C3 bond. The incorporation of the amidic CO into a piperazinone ring facilitates the optimal orientation of this group for chelating Mg²⁺ ions in the active site, and the saturated N,O-heterocycle is important for the inhibitory activity against the Q148K mutated enzyme [159]. Dolutegravir and Bictegravir are formulated as tablets for daily administration, whereas Cabotegravir is formulated as a long-acting injectable suspension to be administered monthly or bi-monthly [160].

Compounds 38–40 have the same central core and are prepared using similar pathways (Scheme 34) [159]. Compounds 38 and 40, developed at GlaxoSmithKline, were prepared from carboxyamide 247 via a reaction with a suitable amino-alcohol to close the hemiaminal ring. Removal of the benzyl protective group by means of catalytic hydrogenation on 248 and 249 then yielded the desired compounds [161]. In the initial syntheses, pyridine 247 was prepared starting with maltol through several synthetic steps (not shown in Scheme 34); later, a faster method was developed starting with methyl 4-methoxy-3-oxobutanoate, which led to the 5-methoxy derivative 250. This intermediate was reacted with (S)-2-aminopropanol, obtaining piperazinone 251. Treatment with 2,4-difluorobenzylamine and CDI gave the corresponding carboxyamide, on which demethylation was accomplished with magnesium bromide, obtaining 40. A similar pathway was also applied to the synthesis of 38 using (R)-3-aminobutanol.

Compound 250 was also used by Gilead in the synthesis of 39 [162]. The reaction with (1R,3S)-3-amino cyclopentan-1-ol gave hemiaminal 252, which was transformed into the final compound 39 using the same method seen before.

4. Role of the Piperazine Moiety

Forty new small molecules carrying a piperazine ring were approved by the FDA between January 2011 and June 2023. Among them, the largest therapeutic class is represented by kinase inhibitors (1, 2, 4, 6–8, 18–23, 29, 31 and 32), developed to treat different types of cancer, even if one of them (Nintedanib, 20) was approved for a different indication (treatment of idiopathic pulmonary fibrosis). The second largest group of piperazine derivatives consists of central nervous system (CNS) receptors modulators (3, 10, 11, 13–17, 26, 33 and 34). The piperazine is indeed a recurrent motif in these two therapeutic classes [4].

The role of the piperazine moiety in compounds 1–40 is various. For some compounds, it is related to the modulation of the physicochemical properties, such as basicity and solubility, which positively affect pharmacokinetics. This has been reported for Entrectinib 8 [47], Bosutinib 18 [93], Ponatinib 19 [96], Brigatinib 23 [106] and Olaparib (27) [120]. Additionally, the protonated piperazine N-atom may contribute to the interaction with the target macromolecule, as is suggested for the kinase inhibitors Palbociclib 1, Ribociclib 2, Abemaciclib 21 [11] and Nintedanib 20 [97] and the RNA slicing modifier Risdiplam 30 [132]. A particular example is Maralixibat 24, whose piperazine ring is included in a DABCO structure; the quaternary N-atom limits absorption, allowing the pharmacological activity to be carried out from the ileal lumen [99]. Moreover, the insertion of DABCO was one of the modifications leading to a crystalline and non-hygroscopic compound [163]. In the case of Venetoclax 12, the piperazine moiety was inserted to increase polarity in specific parts of the molecule in order to limit the interaction with plasma proteins [164].

In some other instances, the insertion of a piperazine moiety has improved safety. For Infigratinib (7), the N-ethylpiperazine group was chosen among other basic groups because it prevented the inhibition of cytochrome P450 isoforms [45]. For Zavegepant (26), replacing (N-methylpiperidinyl)piperidine with (N-methylpiperidinil)piperazine reduced nasal irritation after the administration of the drug [118].

Regarding the CNS receptor modulators 3, 13–17 and 34, which act on dopaminergic and/or serotoninergic systems, the arylpiperazine moiety is part of the pharmacophore. This structural motif allows interaction at the orthosteric site of these receptors, since the N¹-aryl moiety is inserted in a hydrophobic pocked formed by aromatic residues and is located close to the aspartate residue, which establish the pivotal ion–ion interaction with the basic (and protonated) piperazine N⁴-atom [165,166,167].

Some drugs contain the piperazine ring because it was already present in the lead (hit) in the discovery campaign (25 [111], 28 [126], 31 [168] and 32 [142]), and it was conserved in the final molecule. For some other compounds, information about their design and optimization have not yet been reported in the literature, so the role of the piperazine moiety cannot be properly evaluated.

5. Conclusions

The analysis of the methods used to prepare compounds 1–40 shows that most synthetic routes use a synthon that already contains the piperazine ring (compounds 1–13, 15–31). As said in the introduction, many useful N-alkyl, N-acyl or N-protected piperazines are commercially available, often at low costs. The building of the piperazine (compounds 14, 32 and 34–37) or piperazinone ring (6, 38–40) is necessary or suitable for only a few compounds.

Regarding the reaction applied to the synthesis of N-arylpiperazines, Pd-catalyzed Buchwald–Hartwig coupling is more often used in discovery chemistry than in process chemistry due to the concern regarding the use of large quantities of expensive Pd catalysts, whose complete removal from the mixtures has been demonstrated in some cases to be difficult. Safety and environmental concerns regarding the use of Cu catalysts likely also limits the use of the Ullmann–Goldberg reaction. The SN_Ar reaction is most often utilized in process chemistry due to the ease of work-up and the purification of the reaction mixtures. Venetoclax 12 is an exception, because the original SN_Ar reaction is replaced with a Pd-catalyzed amination reaction. In this way, the whole process is improved by reducing the problems connected to the handling of an unstable intermediate (104) and by exploiting a cleaner preparation of a starting compound (109).

To prepare N-alkyl derivatives, the reaction of piperazine reactants with alkyl bromides or chlorides is often helped by the addition of sodium, potassium or ammonium iodide to improve the yield and to avoid the formation of by-products that complicate the purification of the drug. Only in one case, an alkyl mesilate is used (17). When reductive amination is applied, sodium triacethoxyborohydride is employed in the preparation of compounds 12, 17, 21 and 25, and it is formed in situ from sodium cyanoborohydride and acetic acid in the synthesis of 23. It is of note that, in the synthesis of 21, sodium triacethoxyborohydride works well for the reductive amination of 6-bromo-nicotinaldehyde (99% yield), but in the case of the less reactive nicotinic derivative 153, the reaction is not complete, leading to the application of a different method (Leuckart–Wallach reaction, using formic acid).

Ab initio synthesis of the piperazine ring is performed for N-aryl derivatives 14 and 16 starting with an aniline precursor and bis(2-chloroethyl)ammine or iethanolamine, respectively, and for the Adagrasib building block 209, it is performed by condensing dibenzylethanediamine and epichloridrine. The synthesis of 33 and 34 involves the preparation and/or functionalization of piperazin-2-one derivatives, whereas the complex compounds 35–37 require ad hoc methods. Finally, the piperazinone rings of 6 and 38–40 are easily formed via the intramolecular cyclization of amine and ester group intermediates.

In general, synthetic methods are optimized through the careful control of the reaction conditions regarding solvents, reagents, temperatures and work-up procedures.