Heme arginate is still the therapy of choice for patients with acute porphyria attacks . It can also be used prophylactically . The drug is, however, only partially efficacious in the small group of patients with recurrent attacks . Hence, there is a medical need to develop novel drugs based upon enzyme replacement through recombinant enzyme technology, induction of enzyme production by mRNA technology, recovery of enzyme functionality , or inhibition of ∂-ALA-synthase-1 by targeting specifically the mRNA by a synthetic chemically modified double-stranded small interfering ribonucleic acid (siRNA).
Utilizing the latter principle, givosiran is a novel innovative therapy which has been shown to be effective and safe in a phase 3 trial  which led to the registration in the USA and EU. Major adverse effects reported in this trial were renal function impairment in 15%, injection site reactions (25%), and rash (6%) as well as increase of alanine aspartate transferase (8%).
We have treated two female patients with recurrent attacks within the Envision study. After initial excellent clinical response, both patients developed adverse effects (whole body skin rash or pancreatitis, resp.) which unfortunately led to the discontinuation of the therapy with givosiran.
When looking for possible explanations, we unexpectedly observed extreme elevations of the plasma homocysteine levels in both these patients.
Plasma homocysteine elevations in patients with acute porphyrias have first been reported by To-Figueras a decade ago : they found homocysteine elevations up to 80 μmol/l in 24 patients with AIP when compared to healthy controls. In one of these patients, they measured plasma homocysteine before and after repeated heme arginate infusions and observed a reduction by the infusion but a rapid recovery at the initiation of the next session. Although they observed low levels of vitamin B6 in their cohort, they could not correlate normal and lower B6 levels in the patients with hyperhomocysteinemia. Alternatively, they are discussing a reduction of cystathionine-ß-synthase activity which is not only a B6 (pyridoxal 5′-phosphate PLP) but also heme-dependent enzyme .
The studies were recently extended by Ventura and coworkers . They investigated 46 acute porphyria (AP) patients which they divided in three groups: symptomatic patients (AP-SP), patients with biochemical alterations (AP-BA), and asymptomatic carriers (AP-AC). On average, symptomatic patients have higher plasma homocysteine values (27.6 μmol/l with the highest value measured being 80 μmol/l) than AP-BA and AP-AC individuals (17.1 or 10.7 μmol/l, resp.). Since they found lower B6 levels in their cohort as compared to healthy Italian subjects, they hypothesized that the induction of ∂-ALA-synthase-1 may redirect B6 from other metabolic pathways such as B6-dependent homocysteine metabolism and by this cause hyperhomocysteinemia.
When we decided to analyze plasma homocysteine in our patients, we expected normal or slightly elevated levels since both patients had clinically responded well to the treatment with givosiran. However, we observed levels which in both patients exceeded 100 μmol/l.
Severe homocysteinemias (between 100 and 500 μmol/l) are usually not seen in adults but only observed in newborns with inborn cystathionine-ß-synthase (CBS) defects . If left untreated, these patients develop developmental impairment as well as thromboembolic complications (such as stroke), pancreatitis, or skin alterations.
Looking for inborn errors of homocysteine metabolism in our patients, we identified a heterozygous mutation at position c.677 (C > T) of the MTHF-reductase gene in patient A and a homozygous variant in patient B. However, individuals with these polymorphisms usually show plasma homocysteine levels only up to 50 and on very rare occasions up to 100 μmol/l.
NGS sequencing of nine other genes relevant for the homocysteine metabolism (ABCD4, CBS, LMBRD1, MMACHC, MMADHC, MMUT, MTR, MTRR, and PRDX1) did not reveal any additional mutated genes.
Homocysteine is degraded by CBS which in contrast to other enzymes involved in amino acid metabolism not only requires vitamin B6 but, in addition, also heme for its full activity (Fig. 3). CBS not only regulates homocysteine metabolism but also contributes to the biosynthesis of the gaseous transmitter H2S through which it is involved in cellular energetics, redox status, DNA methylation, and protein modification . Moreover, through the transsulfuration pathway, the production of cysteine and the antioxidant glutathione are regulated . In CBS, heme is not directly involved in the catalytic mechanism but guarantees the three-dimensional folding of enzyme mandatory for enzyme activity  and also serves as a signal molecule.
Thus, we hypothesized that givosiran causes an acquired CBS deficiency through depletion of the heme pool which leads to a disturbance of homocysteine degradation aggravated by concomitant hetero- or homozygous MTHFR polymorphisms (Fig. 4).
When we gave heme arginate to both our patients, plasma homocysteine levels rapidly fell but increased again similar to what had been observed previously . Only after disappearance of the activity of givosiran which lasts at least 3 months , they fell to fluctuating values in the moderately elevated range.
We have reported these results to the sponsor of the study as well as all other Envision study centers . Meanwhile, strongly elevated plasma homocysteine levels have also been observed in other porphyria centers . Moreover, in a study on frozen samples from the Envision study patients, Alnylam has reported elevated plasma homocysteine levels, however, without providing any further details .
Impairment of CBS activity may not be the only consequence of heme deficiency but could even be the tip of an iceberg: other heme-dependent metabolic pathways such as tryptophan or testosterone degradation may also be involved . Moreover, in recent years, in addition to the prototypical functions of oxygen metabolism, electron transfer, and CYP450 activities, a plethora of new roles of a mobile (also called free or regulatory) heme in signal transduction has been elucidated . Hence, only metabolomic studies will help to find out which metabolic pathways are being influenced by givosiran treatment .
In addition, side effects of givosiran such as renal impairment may not be a direct effect but may also be caused by disturbed hepatic metabolism with secondary effects on the kidney .
Recently, Nakajima′group in Japan has developed an experimental knockdown model making mice heterozygous for ∂-ALA-synthase 1 . This model resembles the siRNA approach: in their mice, the authors have observed a reduction of free, but not total heme in hepatocytes and also impaired glucose intolerance and insulin resistance .
To which extent severe homocysteinemia (levels above 100 μmol/l) has caused or contributed to the adverse effects observed in our two patients cannot be decided at present since such high values occur only in newborns with inborn homocystinuria. Hence, most studies in adults are being carried out on patients with intermediate (31–100) or moderate (16–30) plasma homocysteine levels. However, pancreatitis is a common feature in patients with inborn homocystinuria [18, 32, 33]. Moreover, hyperhomocysteinemia has been associated in several studies with acute and chronic pancreatitis [34,35,36,37]. Alternatively, olanzapine taken by our patient (B) can also cause pancreatitis . Our patient, however, was on a low dose without any problems for 3 years and later reexposed to the drug without exacerbation of the pancreatitis.
Givosiran can cause local injection reactions and rash as reported in the Envision study . Excessive homocysteinemia may aggravate these side effects through the induction of endothelial dysfunction in the skin [39, 40] and mitochondrial dysfunction  in the liver (see above). Part of these effects might be due to a N-homocysteinylation of various proteins [42, 43].
Interestingly enough, fibrosis of the intestinal tract—as seen in patient B—is promoted by homocysteine in an experimental animal model .
From our observations, we conclude that givosiran aggravates the disturbance of homocysteine metabolism which is already present in many patients with acute porphyrias. Mechanism of action may be a reduction of free heme in the hepatocyte. The degree of further homocysteine elevation will be determined (provided that renal function and vitamin status are normal) by concomitantly present polymorphisms of the MTHFR gene. Since givosiran therapy is a continuous therapy, chronic severe homocysteinemia may lead to various long-term complications.
In order to avoid these, we recommend that all patients who are considered for therapy with givosiran should be assayed for plasma homocysteine levels prior to initiation of therapy and while being on therapy. In case of pretherapeutical homocysteinemia, patients should be tested for vitamin B6, folic acid, and vitamin B12 levels as well as for MTHFR polymorphisms. Depending upon the results, vitamin replacement therapy (possibly also with betaine) should be implemented.
Further studies will be needed to reveal whether the current monthly treatment with givosiran has to be modified.