In the first half of the XXieth century, Hans Fischer, the German Nobel Prize in Chemistry, demonstrated that haem precursors, porphyrins, ‘are compounds which make grass green and blood red’. Although haem is synthesized in every human cell for respiratory and oxido-reduction reactions, it is mostly produced in the erythropoietic cells for haemoglobin synthesis and in the liver parenchymal cells for cytochrome and other haemoprotein synthesis. In the last 40 years, porphyrias, a group of eight inherited metabolic disorders of the haem biosynthesis pathway, were characterized (Fig. 1) and the liver appeared as a central organ in the pathophysiology of porphyrias. Each type of porphyria is the result of a specific enzyme alteration in the pathway. Specific patterns of accumulation of haem precursors, δ-aminolevulinic acid (ALA), porphobilinogen (PBG) and porphyrins, are associated with characteristic clinical features: acute neurovisceral attacks, skin lesions or both [1, 2]. Acute hepatic porphyrias (AHP) are a group of three autosomal dominant disorders with very low clinical penetrance: acute intermittent porphyria (AIP), variegate porphyria (VP) and hereditary coproporphyria (HCP). The liver plays a major role in the pathophysiology of these porphyrias. The majority (>80%) of the relatives within affected families who inherit an autosomal dominant porphyria trait remains asymptomatic throughout their life (so-called latent or presymptomatic subject). The misunderstanding of porphyrias well defined by Professor Anthony Mc Donagh (UCSF, 1997) as ‘obscure diseases with confusing names considered only when the need for a diagnosis is desperate’ associated with the low clinical penetrance make porphyrias usually under-diagnosed in the general population.


Figure 1. Regulation of haem biosynthesis in the liver and better health care for patients. AHP, Acute hepatic porphyrias; AIP, acute intermittent porphyria; HCP, hereditary coproporphyria; VP, variegate porphyria; ALAS, δ aminolevulinic acid synthase; ALAD, ALA dehydratase; PBGD, porphobilinogen deaminase; UROS, uroporphyrinogen III synthase; UROD, uroporphyrinogen decarboxylase; CPO, coproporphyrinogen oxidase; PPOX, protoporphyrinogen oxidase; FECH, ferrochelatase; HO, haem oxygenase; TRP, tryptophan; NO, nitric oxide; CO, carbon monoxide.

Download figure to PowerPoint

The association between hepatocellular carcinoma (HCC) and hepatic porphyrias has been first invoked in a nonacute hepatic porphyria: porphyria cutanea tarda (PCT). However, in this context, both HCC and PCT appeared as a consequence of an underlying liver disease such as chronic alcohol abuse, hepatitis C and B, HIV or hemochromatosis, each contributing to fibrosis and cirrhosis. The association between cirrhosis and HCC has long been recognized and it is well established that cirrhosis is the greatest risk factor for the outcome of HCC. Nowadays, the diagnosis of PCT is easy to handle and patients benefit of symptomatic and aetiopathogenic treatments that limit hepatic injury and the progression to cirrhosis. Therefore, the occurrence of HCC is dramatically decreased, and HCC is no longer considered as a common feature in PCT [1].

More recently, the finding of a significant association between AHP and HCC, mainly without underlying liver disease, became an important issue. HCC is the most common primary malignant tumour in adult liver and between 10% and 40% of HCC arise in noncirrhotic livers [2]. In the 1980s, two retrospective studies and one case–control study from Scandinavian countries were the first to suggest an association between primary liver cancer and acute hepatic porphyrias [3–5]. In 1998, a prospective cohort study assessed the crude annual incidence rate of primary liver cancer in 650 French AHP gene carriers followed for 7 years [6]. Primary liver cancer was found mainly in AIP subjects but also in VP and HCP, both in symptomatic and latent carriers [7–9]. Most patients did not have any HCC aetiologic factors and histological study of the nontumoural liver tissue was usually normal. There was no evidence of fibrosis, inflammatory-cell infiltrate, necrosis nor of iron overload. Altogether, these studies, associated with the evidence that acute hepatic porphyria precedes HCC in patients with porphyria, clearly suggest that AHP should be considered as a cause of HCC mainly when it occurs in the absence of underlying liver disease.

Although most patients with acute porphyria have a normal life, Kauppinen et al. [10] reported that, in Finland, 10% of patients with AIP died from hepatoma and suggested that yearly ultrasound examination of the liver and measurement of serum α-foetoprotein should be undertaken in individuals older than age 50 years. However, the potential benefit of this follow-up was not evaluated.

The study by Innala E and Andersson C [11] in the present issue of the Journal of Internal Medicine brings an outstanding and substantial contribution towards improvement of health care and management of AHP families. First, the authors carefully estimated the annual incidence rate of HCC in patients with AIP above 50 years of age prospectively followed during a long 15-year period. The 0.8% annual incidence of HCC in AIP gene carriers in this Swedish population clearly demonstrates the relationship between HCC and AIP. As pointed out by Innala et al., the over representation of the founder mutation W198X of the HMBS gene in Sweden might limit the generalization of their conclusions to other countries. However, in Switzerland, and even more so in France where private mutations affect AIP families, the annual incidence of HCC was found to be 0.3% and 0.16%, respectively [7, 9]. Secondly, the absence of liver cirrhosis associated with HCC was confirmed by Innala et al., and interestingly, 27% of AIP gene carriers who developed HCC were asymptomatic. Additionally, this study pointed out the usefulness of a systematic screening for a latent AHP gene carriage in patients newly diagnosed with HCC without cirrhosis.

Despite a marked improvement in management of HCC, liver cancer carries a dismal prognosis that is partly explained by the late diagnosis and partly by the clinical context of the patient, both compromising potential therapeutic strategies [12]. The particular strength of this study is to demonstrate the benefit of annual screening using liver imaging for AIP gene carriers over 50 years of age. HCC is primarily a disease of older men. The incidence of HCC generally increases with age. In Western Europe and USA, most patients with HCC are between 50 and 75 years of age, although there are geographic differences. Without underlying chronic liver disease, clinical manifestations of HCC are very limited, thereby contributing to a late diagnosis. Because of the dismal prognosis of the disease, increasing emphasis has been placed on detecting early, asymptomatic HCCs that are at a potentially curable stage. A more favourable prognosis is reported amongst patients with well-compensated or no cirrhosis, good performance status and small tumours of <3 cm diameter diagnosed during patient follow-up [13]. Altogether, it is not surprising that a screening programme appears beneficial in a cohort of AIP patient at increased risk of developing HCC on noncirrhotic liver. Furthermore, the age of 50 appears appropriate to initiate the follow-up.

In HCC, routine liver function tests are variably abnormal, often reflecting the underlying cirrhosis without consistent pattern of alterations. Serum α-foetoprotein is raised in most symptomatic tumours but small cancers are associated with lesser or even normal levels. Thus, serum α-foetoprotein is not considered as a reliable diagnostic test for screening patients with cirrhosis [14] and did not appear as a useful screening tool in the study by Innala et al. Ultrasonography is a sensitive technique for identifying nodules larger than 1.5 cm and should thus been used for early detection, whereas magnetic resonance imaging is a more accurate procedure for characterizing tumour masses. The screening interval is still in debate and the annual screening should be proposed [14].

Recent studies demonstrated that several risk factors for metabolic syndrome, especially diabetes and obesity, were found in a substantial percentage of patients with cryptogenic cirrhosis [15]. Unfortunately, the natural history of HCC and the mechanisms of liver damage in metabolic disorders of the heme biosynthetic pathway are poorly understood. Various hypotheses that are not mutually exclusive have been proposed. The leading hypothesis is that the pro-oxidant haem precursor ALA (and/or PBG) overproduced by the liver is toxic [16]. Conversely, formation of haemoproteins may be compromised because of the inherited enzyme deficiency. Interestingly, in hereditary type I tyrosinemia, the risk of HCC is very high with only 40% of patients surviving beyond 2 years after developing HCC [17]. Fumarylacetoacetate, the intermediate metabolite that accumulates in type I tyrosinemia, inhibits haem biosynthesis pathway in the liver and leads to ALA accumulation. These data support the putative pro-oxidant and carcinogenic role of ALA in the development of HCC both in tyrosinemia and AHP.

In conclusion, Innala E and Andersson C masterfully demonstrated in their study the significant relationship between HCC without cirrhosis and AIP. The increasing knowledge in the molecular pathogenesis of HCC creates an encouraging trend in the management of HCC without cirrhosis. This opens the era of new therapeutic trials for AIP and all other AHP patients newly diagnosed with HCC without underlying disease (Fig. 1). Sweden and other Northern European countries, well organized for epidemiological studies and for taking care of rare disorders, mainly contribute to these improvements by the follow-up of the aging population of patients with porphyria, as perfectly illustrated by this remarkable study. It is also the first demonstration of the benefit of an annual liver screening by ultrasonography of all AIP gene carriers (symptomatic or latent) aged >50 years (Fig. 1). This follow-up should be extended to all AHP gene carriers (AIP, VP, HCP) outside Sweden. It should be used as the referent study to develop a common approach to prevent HCC in acute porphyrias. This study strongly supports the guidelines of the European Porphyria Network (EPNET; and reinforces the necessity of a collaborative network between porphyria centres to provide better health care for patients and their families.

Conflict of interest statement

  1. Top of page
  2. Conflict of interest statement
  3. References

No conflict of interest was declared.


  1. Top of page
  2. Conflict of interest statement
  3. References