There is much observational evidence that exposure to organic solvents, paints, glues, fuels, motor exhausts and other environmental hydrocarbon contaminants may induce glomerulonephritis and also worsen renal function in a large number of patients 123. Indeed, this hypothesis satisfies all except one of Hill's criteria for causality 3. In spite of that the significance of hydrocarbon exposure has not been generally acknowledged and most current textbooks mention little if anything about this issue. Arguments often used by sceptics are that no probable mechanisms are known, that kidneys from animals exposed to hydrocarbons have shown tubular damage mainly, and that no specific glomerular pattern of hydrocarbon-associated glomerulonephritis has been identified in human beings. Glomerulonephritis has indeed been produced in a few experiments by exposing animals to hydrocarbons 3. They are little known, and as no review of this subject has been published previously I found it relevant to perform a systematic search for such studies and found twenty-six.

Using Medline and Toxnet I sought experiments that had produced glomerulonephritis by exposing animals to hydrocarbons. The search strategy included the formula (glomerulonephritis OR glomerulopathy) AND experiment* AND (hydrocarbon* OR solvent* OR X) where X was substituted by a large number of various hydrocarbons with putative toxicity and commonly used in the industry or elsewhere. Relevant papers were also sought in the reference lists of the studies. Papers that mentioned glomerular changes of any kind in the abstract were required as were papers without an abstract. All papers in the Western European languages were considered and included if appropriate.

Twenty-six experiments were identified, where the authors had noted microscopical changes in the kidneys of the animals similar to those seen in human glomerulonephritis after having exposed them to various hydrocarbons 45678910111213141516171819202122232425262728. One experiment was reported in two papers 45, two groups used two different hydrocarbon 723; totally 13 different hydrocarbons were used in 26 experiments. In two experiments 915 the animals were exposed to a single dose of the hydrocarbon, in the rest they were exposed intermittently. In 15 experiments, mainly the more recent ones [69111213141617182224252627, 35], unexposed control animals or animals exposed to neutral substances were included. In all of them the renal changes, if any, were mild and did not exceed those seen in normal, aging rats. In ten experiments, the kidneys were examined by light microscopy (LM) only, in the rest by immunofluorescence microscopy (IM), and/or scanning or transmission electron microscopy (EM) also. Glomerular proteinuria was found in 10/12 experiments that included an examination of the urine. Evidence of renal insufficiency was found in 5/8 experiments that included a determination of renal function. In 19 studies the tubulointerstitial tissue were described also and in all of them varying degrees of damage were noted, in particular the two experiments that produced anti-TBM nephritis1920. Findings similar to most of the human subgroups were seen, including IgA nephritis 18, mesangial 1316, crescentic 614, proliferative 620 and focal segmental proliferative 1920 glomerulonephritis, focal 18, focal segmental 172227 and total 1126 glomerular sclerosis, minimal change nephropathy 112025, anti-TBM nephritis 19 and glomerulonephritis associated with periarteritis nodosa 9. A summary of the findings is given in Additional file 1.

As shown in this review glomerular changes similar to human glomerulonephritis have been found in 26 expeiments by exposing animals to hydrocarbons, a strong support to the view, that such exposure may induce glomerulonephritis in human beings. The histological documentation in the early studies was cursory, but the more recent experiments, where the authors had examined the renal tissue by EM and/or IM also, have mimicked most of the human subgroups successfully. Few models of experimental glomerulonephritis using manipulation of the immune system have resulted in more harm to the kidneys than trace or transient proteinuria, unless they have included the use of Freund's adjuvant 29, the main ingredient of which is a mixture of hydrocarbon oils. In contrast, most of the experiments reviewed here, that included laboratory measurements, resulted in severe proteinuria and renal failure.

Up to five different subgroups of glomerulonephritis were seen after exposure to the same hydrocarbon. In some of the models it was obvious that they represented various stages of the disease. Genetic differences may also play a role, as many different animals were used. The most likely explanation is probably that the hydrocarbons used may not have been pure. Volatile hydrocarbons are purified by distillation, but as raw oil is composed of thousands of different hydrocarbons, many of which have a similar boiling point it is difficult to obtain a pure hydrocarbon. Therefore it is not possible to know whether it is the hydrocarbon, the name of which is printed on the bottle, that has resulted in a certain subgroup, or whether it has been caused by one of the impurities.

One of the arguments against the idea that hydrocarbon exposure may cause glomerulonephritis has been that a specific glomerular pattern has never been identified, but as almost all subgroups were found in these experiments this argument is obviously untenable.

The experiments offer several explanations regarding the mechanisms behind hydrocarbon-induced glomerulonephritis. Each of them may be sufficient by itself, although a complex co-operation seems more likely.

By combining with renal proteins hydrocarbons may act as haptens and induce autoimmunity against kidney cells. This mechanism was proposed by Nakajima et al in experimental DNCB nephritis 20 and may also have been operating in the experiment by Hartmann et al. 9.

A causal role of glomerular immune complex formation was unlikely in most of the studies because glomerular deposits of immunoglobulin and complement appeared late in the course, if at all and may therefore have been secondary phenomena without pathogenetic importance. This is in agreement with many clinical observations, where deposits of Ig and C in the glomeruli have been seen without evidence of renal disease 29. Also, no study has ever found an association between degree of immune complex deposition or degree of glomerular damage, and the clinical course or outcome 2930.

Hydrocarbons may influence T-cell functions leading for instance to an extrarenal production of cytokines with harmful effects on the glomerular epithelial and/or endothelial cells as in human minimal change nephropathy. This mechanism was likely in the experiment with maleic vinyl ether anhydride (MVE-2), diacetylbenzidine and carbon tetrachloride, as both podocyte fusion and proteinuria was seen immediately after the exposure 1115252627 whereas immune deposits, if any, were seen much later 26. In favour of an effect directed against extrarenal T-cells were the findings that pre-treatment with irradiation 2526 or methylprednisolone 25 abrogated proteinuria, and also that radiolabeled MVE-2 was found to be located mainly to the reticuloendothelial system, not to the kidneys themselves 31.

The effects of hydrocarbons on the immune system are probably more complicated. Many hydrocarbons are able to suppress a wide range of immunological functions including the lymphoproliferative response to T-cell mitogens, memory cell and macrophage function, delayed hypersensitivity, serum immunoglobulin synthesis and susceptibility to infectious challenges, both in experimental animals and in man 32.

All experiments resulted in various degrees of tubular damage. Such damage may lead to an obstruction of the tubular flow, an increased postglomerular resistance and eventually a decreased glomerular filtration rate 33. Together with an increased glomerular permeability these effects may result in a secondary trapping of circulating macromolecules. That proximal tubular damage may play a role in the causation of glomerulonephritis is suggestive also from the fact that most nephritogenic chemicals, for instance mercury, gold, silicium and lithium, are tubulotoxic as well, and that a large number of studies have found that the renal function and the course in glomerulonephritis is strongly predicted by the degree of tubulointerstitial damage, but not by the degree of glomerular damage 29.

The finding in one of the experiments that buffalo rats were susceptible to exposure, but not mice 7, and in another that female rats were more susceptible than male rats 6 confirms the experience from clinical studies that a genetic or sexual predisposition is necessary. This finding may also explain why many similar experiments have failed to produce glomerulonephritis.

The animal experiments are in accordance with the results from many cross-sectional and case-control studies. Thus, in 14 studies of workers exposed regularly to hydrocarbons significantly more had urinary findings typical of early glomerulonephritis than had age and sex-matched control individuals. The findings included a pathological urinary sediment, increased excretion of albumin and/or transferrin and/or glomerular antigens, and in three of them renal failure was also more prevalent 3. And in a meta-analysis of 18 case-control studies odds ratio for exposure was 1.1 or lower in four studies that included early glomerulonephritis only, whereas in all studies that included patients with failure the ratio was 1.7 or, more often, much higher, in particular in studies that included patients with end-stage renal failure only 3.

The successful imitation of human glomerulonephritis achieved in 26 experiments by exposing animals to various hydrocarbons is a further confirmation, suggested by numerous observational studies, that such exposure may induce glomerulonephritis in predisposed individuals. To convert this knowledge to benefit for the patients demands more research, however. Most important would be prospective studies of the effect of a discontinuation of the exposure. If the course in patients with renal failure could be reversed in this way, as was the case in a few retrospective studies and in a small prospective one 3, it would be the final proof that hydrocarbon exposure is causal in glomerulonephritis. At the same time it may become a major improvement in the prevention of terminal renal failure because hydrocarbon exposure is prevalent in patients with renal failure; in the four case-control studies that included patients with end-stage renal failure only, more than 50 % reported about significant exposure. As the microscopic findings obviously cannot be used as a marker for hydrocarbon exposure, such studies demand a thorough questioning of all patients with glomerulonephritis for possible environmental hazards, preferably in co-operation with experts in occupational medicine.

The author(s) declare that they have no competing interests.