Short title: Evening primrose oil for uremic pruritus

All correspondence should be addressed to: Hirotoshi Echizen, M.D.& Ph.D., Department of Pharmacotherapy, Meiji College of Pharmacy, 1-22-1 Yato- cho, Tanashi-city, Tokyo 188-0001, Japan Tel. & Fax. +0424-21-0281, e-mail:echizen@my-pharm.ac.jp

Key words: essential fatty acids, evening primrose oil, γ-linolenic acid, linoleic acid, hemodialysis, uremic skin symptoms, randomized double-blind trial.

【 Abstract 】

Abnormalities in plasma composition of essential fatty acids may be associated with the etiology of pruritus and other skin problems in patients undergoing hemodialysis.To study whether an oral supplementation with omega-6 (n-6) essential fatty acids (EFAs) would restore deranged plasma essential fatty acids and ameliorate skin symptoms, 9 and 7 dialysis patients were randomly assigned to receive eitherγ-linolenic acid (GLA)-rich evening primrose oil (EPO) or linoleic acid (LA) (2g/day each) for 6 weeks. Plasma concentrations of EFA were analyzed by gas chromatography and uremic skin symptoms were assessed for dryness, pruritus and erythema by questionnaire and visual inspection in a double-blind manner. The patients given EPO exhibited a significant (p < 0.05) increase in plasma dihomo-γ-linolenic acid (DGLA) (a precursor of anti-inflammatory prostaglandin E1) with no concomitant change in plasma arachidonic acid (AA)(a precursor of pro-inflammatory prostaglandin E2 and leukotriene B4).In contrast, those given LA exhibited a significant (p < 0.05) increase in LA but not in any other n-6 EFAs, whereas they exhibited a significant (p <0.05) decrease in plasma docosahexaenoic acid (DHA).The patients given EPO showed a significant (p < 0.05) improvement in the skin scores for the 3 different uremic skin symptoms over the baseline values and a trend toward a greater improvement (0.05 < p <0.1) in pruritus scores than those given LA.Results indicate that GLA-rich EPO would be a more favorable supplemental source than LA in terms of shifting eicosanoid metabolism toward a less inflammation status through modifying plasma concentrations of their precursor n-6 EFAs. Further studies are required to confirm the efficacy and safety of EPO therapy for the treatment of uremic pruritus.

【 Introduction 】

Uremic pruritus is one of the most disabling clinical symptoms in patients with end-stage renal disease (ESRD) undergoing maintenance hemodialysis [for review, see ref.1]. As many as 80% of hemodialysis patients reportedly are suffering from uremic pruritus one time or another [2]. The pruritic skin may appear normal except for dryness, but is often complicated by inflammatory lesions (e.g., lichen simplex, prurigo nodularis) secondary to scratching [3]. While dryness of the skin [4], secondary hyperparathyroidism with altered plasma divalent ion concentrations (e.g, elevated calcium-phosphorus product) [5], proliferation of cutaneous mast cells with elevated plasma histamine levels [6,7] or a combination thereof have been implicated in the pathogenesis of uremic dermatoses, their precise cause(s) remains largely obscure. Thus, numerous therapeutic attempts made so far based upon the above-described speculative etiologies have been proven only partially or temporarily effective [1-3].

Various lines of evidence suggest that abnormal plasma and possibly cutaneous composition of essential fatty acids (EFAs) in patients with ESRD may play an important role in the development or aggravation of uremic skin symptoms. Deficiency of EFAs was shown to produce skin symptoms (i.e., redness, scaling and itching) resembling uremic dermatoses in rats [8] and humans [9]. In addition, previous studies [10,11] have revealed that patients with ESRD undergoing hemodialysis had significantly reduced serum concentrations of several EFAs [i.e., linoleic acid (LA), a-linolenic acid and arachidonic acid (AA)] as compared with the healthy subjects.Because LA is a vital constituent of the water-impermeable layer in the epidermis [12], depletion of plasma LA may lead to xerosis of the skin. On the other hand, certain omega-6 (n-6) EFAs that are converted from LA may be involved in the inflammatory changes of uremic dermatoses via modulation of eicosanoid metabolism. While AA is converted to a pro-inflammatory eicosanoids (e.g., prostaglandin E2) known to enhance itching [13], γ-linolenic acid (GLA) is converted to a less inflammatory eicosanoid, prostaglandin E1, via dihomo-γ-linolenic acid (DGLA) [14] (Fig.1).In addition, another hydroxylated metabolite of DGLA, 15-hydroxy DGLA, is a potent inhibitor of 5- and 12-lipoxygenases that produce pro-inflammatory leukotrienes (e.g., leukotriene B4) [15]. Thus, there is a possibility that supplementation with LA or GLA might restore deranged plasma and possibly cutaneous compositions of EFAs and thereby ameliorate uremic skin symptoms.

Previous studies have shown that the conversion of LA to GLA (i.e., Δ-6-desaturation) is a rate-limiting step in the n-6 EFA metabolism in humans [16]. At present it remains unknown whether LA or GLA would be better as an oral supplemental source of n-6 EFAs and how much of them should be given to produce desirable changes in plasma EFA concentrations in patients undergoing hemodialysis. Here, we present a prospective, randomized and double-blind clinical trial where we compared the effects of 6-week oral supplementation with LA or GLA-rich EPO not only on plasma composition of EFAs but also on uremic skin symptoms in 16 hemodialysis patients.

【 Methods 】

Study design and patients Sixteen patients (7 males and 9 females; aged from 23 to 79 years) undergoing hemodialysis who suffered from pruritus and other uremic skin symptoms (i.e., dryness or erythema) participated in the study. Their demographic, clinically relevant laboratory data and skin symptom scores are shown in Table 1.The adequacy of dialysis therapy was assessed based upon KT/V value that was calculated according to the following formula [17]: KT/V = - ln (R - 0.03) + (4 - 3.5 x R) x UF/BW where R is the post/pre blood urea nitrogen (BUN) ratio, UF is the volume of ultrafiltrate removed (liters) and BW is the postdialysis body weight (kg). Currently, a KT/V value of 1.2 per treatment is considered minimal standard for adequacy [17]. The patients were randomly assigned into two study groups; one was given GLA-rich evening primrose oil (EPO) and the other was given LA. Both groups were given the respective treatments for 6 weeks in a double-blind manner. EPO was provided in capsules containing 360 mg of LA, 50 mg of oleic acid and 45 mg of GLA (EfamolR), and LA was provided in capsules containing 500 mg of LA only. The EPO and LA capsules were visibly indistinguishable each other, and both were kindly supplied by Efamol Ltd., Surrey, UK. Each patient was given two capsules of either EPO or LA twice daily throughout the study period. Following the double-blind phase of the study, those who had been assigned to the EPO group were kept on the EPO treatment for additional 6 weeks to study the biochemical effects of EPO. All patients were kept on a low-potassium and low-phosphorus food throughout the study period under the regular guidance of one of the authors (K. Y-F.). For all study patients hemodialysis was performed with a hollow fiber dialyzer equipped with a cuprophane membrane. Written informed consent was obtained from each patient after thoroughly explaining the purpose of the study. The study protocol was approved by the institutional review board before the study began.

【 Scoring for uremic skin symptoms 】

The severity of uremic skin symptoms (i.e., pruritus, erythema and dryness) was measured by a grading-scale system modified from those employed in previous studies [18,19]. The intensity of pruritus was assessed with a self-administered questionnaire using statements in Japanese which mean as follows: 1 (I rarely itch), 2 (I occasionally itch with mild annoyance), 3 (I itch often and feel mild annoyance), 4 (I itch often. It may be severe and interfere with both rest and activity) and 5 (I itch most of the time. It is severe and interferes with both rest and activity considerably). Dryness of the skin and erythematous lesions were assessed separately by visual inspection by a single investigator (Y.K.) who was unaware of the treatment assignments of each patient: severity was scored as 1 (none), 2 (mild), 3 (moderate), 4 (marked) and 5 (severe). Assessments of skin symptoms were performed at the beginning of the study and weekly thereafter throughout the 6-week treatment period. An initial score of 5 was a prerequisite for study participation.

【 Fatty acid measurements 】

Blood samples for fatty acid measurements were obtained from each patient at the baseline and at week 6 of treatment. For those who were kept receiving EPO for additional 6 weeks in the open phase of the study, blood samples were also obtained at week 12 of the treatment. Blood was drawn from patients before dialysis session. Immediately after collection of venous blood (10 ml) into a heparinized glass tube, the sample was centrifuged at 400 g for 10 min, and separated plasma was stored frozen at - 40 oC until analyzed. Frozen plasma samples were flown packed in dry ice to the Clinical Service, Scotia Pharmaceutical Ltd., Surrey, U.K. and analyzed for n-3 and n-6 series of EFAs as shown in Fig.1 and for some other nonessential fatty acids as shown in Table 2 according to the method reported elsewhere [20]. Briefly, the samples were extracted with chloroform/methanol (2/1, v/v) and the extract was filtered through sodium sulphate, evaporated to dryness and reconstituted in 0.5 ml chloroform/methanol solution. Lipid fractions were separated by thin layer chromatography on silica gel plates (E. Merck, Darmstadt, Germany). The phospholipid, triglyceride and cholesterol ester fractions were methylated separately using boron trifluoride/methanol. The resulting methyl esters of fatty acids obtained from the 3 different plasma lipid fractions were separated and measured by a Hewlett-Packard 5880 gas chromatograph with a 6-foot column packed with 10% silar on chromatosorb WAW 106/230. Helium was used as the carrier gas at a rate of 30 ml/min. Oven temperature was programmed to rise from 165 to 190 oC at 2 oC/min. Detector and injector temperatures were set at 220 oC and 200 oC, respectively. Retention times and peak areas were automatically recorded by a Hewlett-Packard Level 4 integrator. Peaks were identified by comparison with authentic standards from NuChek Prep. Inc., Elysian, MN, USA. Within-assay and between-assay coefficients of variation were < 5%. All analyses were performed by the same investigator (J.S.) who was unaware of the assignments of the treatment for each sample. Data for 15 fatty acids were expressed as mg/100mg lipid for the respective plasma lipid fractions (i.e., phospholipid, cholesterol ester and triglyceride).

【 Other routine laboratory measurements 】

Routine blood chemistry and serum electrolytes were determined at the Department of Clinical Chemistry, Biomedical Laboratories (BML, Kawagoe, Japan) by use of an automatic analyzer (SMAC I, Technicon, Tarrytown, NY, USA). Serum levels of parathyroid hormone were determined by a sensitive double-antibody method in BML.

【 Statistical analysis 】

Demographic and laboratory data obtained from patients who were assigned to either EPO or LA treatment at baseline were compared by Fisher's exact probability test for categorical variables (i.e., gender), Student's t-test for unpaired data for normally distributed variables (e.g., age and laboratory data) and Mann-Whitney U-test for ordered scores (i.e., skin scores). For assessing the effects of EPO and LA on uremic skin symptoms, changes observed in the individual skin scores as compared to the respective baseline values in each patient at weekly clinical inspections were calculated, then the medians and 25 and 75% interquartile ranges were calculated for the different categories of skin scores. Multiple comparisons for the changes in the skin scores throughout the study period were made with the Friedman test for repeated measurements of ordinal scale.When a statistically significant change was observed for overall changes in skin scores, a pair-wise comparison was made between the baseline value and that observed at week 6 by the Wilcoxon signed-rank test. In addition, comparisons were made between EPO and LA groups regarding scores for the 3 different uremic skin symptoms based upon the sum of the scores obtained from week 5 and 6.

Multiple comparisons for the mean values of the plasma concentrations of the respective fatty acids obtained from the EPO group at baseline, 6 and 12 weeks of the treatment were made using analysis of variance (ANOVA) for the repeated measurements followed by Student's t-test for paired data with Bonferroni's correction. Those obtained from patients given LA at baseline and at 6 weeks were compared using Student's t-test for paired data. A p value of < 0.05 was considered statistically significant. Correlations between clinical skin scores for uremic dermatoses and plasma concentrations of EFAs (e.g., DGLA) were analyzed by Spearman's rank correlation. All statistical analyses were performed by SPSS programs (SPSS, Inc., Chicago, IL, USA). Data are reported as either medians and interquartile values or mean values + SD where appropriate.

【 Results 】

Demographic and other clinical data

At baseline no significant differences were observed for any demographic or laboratory data except for gender distribution: the male/female ratio for the EPO group was less (p < 0.05) than that of the LA group (Table 1). No significant difference was observed between the mean KT/V value obtained from patients given EPO and LA. In addition, no significant difference was observed in the pretreatment skin scores between the EPO and LA groups.

Plasma lipid profiles and concentrations of EFAs

Table 2 shows the mean (+ SD) values for the plasma concentrations (mg/100 mg lipid) of EFAs in 3 different plasma lipid fractions obtained from patients before and after the administration of either EPO or LA. No significant differences were observed in any of the plasma concentrations of EFAs between the EPO and LA groups at baseline (Table 2) across the 3 lipid fractions examined. In addition, no significant differences were observed in the gross plasma lipid profiles (i.e., total cholesterol and HDL-cholesterol) between the EPO and LA groups (Table 1) at baseline.

The patients given EPO showed significant (p < 0.01) increases in the mean plasma concentration of DGLA (20:3n-6) in all the lipid fractions at 6 and 12 weeks of the treatment as compared with the corresponding baseline value (Table 2). In addition, they showed significant (p < 0.01 or 0.05) increases in the plasma concentrations of GLA in both the cholesterol ester and triglyceride fractions at 6 or 12 weeks. In contrast, those given LA showed significant (p < 0.05) increases in the mean plasma concentration of LA in the phospholipid and triglyceride fractions at 6 weeks as compared with the corresponding baseline values. However, they showed no significant changes in the plasma concentration of DGLA throughout the study period (Table 2). Furthermore, the administration of LA significantly (p < 0.05 or 0.01) reduced the mean concentrations of docosahexaenoic acid (DHA) in the phospholipid and cholesterol ester fractions at 6 weeks as compared to the baseline value. No significant differences were observed in the gross lipid profiles (total and HDL cholesterol) between the baseline and post-EPO or -LA treatment period (data are not shown).

【 Changes in skin symptom scores 】

As described earlier, no significant differences were observed for the total skin symptom scores (Table 1) or any of the individual symptom scores (data are not shown) at baseline between the EPO and LA groups. There was a statistically significant (p < 0.05) overall improvement throughout the treatment period in the skin scores for the three different categories of uremic dermatoses in the patients given EPO (Fig. 2), whereas no significant changes were observed in the corresponding values in those given LA. The pruritus scores obtained from the EPO group showed a trend toward a greater improvement (0.05 < p < 0.1) than that obtained from the LA group. At baseline no significant correlations were observed between 3 categories of uremic skin symptom scores obtained from overall study patients and any of the concentrations of EFAs examined (data are not shown). No significant correlations were observed between changes in the skin scores and those in the concentrations of any EFAs examined (data are not shown). All patients tolerated well the administration of EPO or LA and none terminated the study prematurely. No adverse clinical symptoms, which might be attributable to the administration of EPO or LA, were observed.

【 Discussion 】

Accumulated data imply that an apparent depletion of n-6 series EFAs (e.g., LA and AA) in patients with ESRD [15,16] might play a role in the development and/or aggravation of uremic pruritus and related skin lesions. However, no attempts have been made to address the question as to whether oral supplementation with a oil product rich in n-6 EFAs would correct deranged plasma fatty acids and ameliorate uremic skin symptoms in patients undergoing dialysis. To our knowledge, this is the first study to evaluate the effects of the oral supplementation with GLA-rich EPO and LA on plasma concentrations of EFAs and uremic skin symptoms in patients on hemodialysis. We have found that the oral supplementation with EPO containing 0.18 g/day of GLA produced significant (p < 0.01) elevations in the plasma concentrations of DGLA in all 3 lipid fractions and a significant elevation in the concentration of GLA in the cholesterol ester and triglyceride fractions at 6 or 12 weeks of the treatment (Table 2). In contrast, supplementation with LA at a dose of 2 g/day significantly (p < 0.05) increased the plasma concentration of LA in the phospholipid and triglyceride fractions but produced discernible changes in neither DGLA nor further n-6 metabolites of EFAs in any plasma lipid fractions examined (Table 2). This finding is in good agreement with previous data obtained from healthy subjects [21] and patients with atopic eczema [22] and suggests that the conversion of LA to GLA (i.e., delta-6-desaturation) is the rate-limiting step in humans (Table 2). Thus, supplementation of EPO (rich in GLA), rather than LA, is an effective means for accelerating the synthesis of DGLA and possibly its further less inflammatory eicosanoid metabolite (i.e., PGE1) in patients with ESRD. In our study the effects of oral supplementation with EPO or LA on the concentrations of n-3 and n-6 EFAs obtained from the phospholipid fraction are largely similar to those obtained from the cholesterol ester fraction, whereas changes in the triglyceride fraction were less evident (Table 2). It is known that plasma concentrations of EFAs in the phospholipid fraction alter less with food intake acutely as compared with those in other lipid fractions [23].

It is worth noting that the supplementation with LA produced a significant (p < 0.05 or 0.01) reduction in the plasma concentrations of DHA in the phospholipid and cholesterol ester fractions, whereas supplementation with EPO did not (Table 2). These findings may be explained by the fact that n-3 and n-6 EFAs are mutual competitive inhibitors at the rate-limiting steps (i.e., 5 or 6-desaturation) in their metabolic cascades [27](Fig. 1). Because LA is a substrate for the 6-desaturase which also metabolizes a-linolenic acid, the administration of LA (2.0 g/day) to patients may have interfered with the conversion of a-linolenic acid to stearidonic acid, thereby reducing DHA levels. In contrast, because the GLA-rich EPO employed in the present study contained a smaller amount of LA (i.e., 1.44 g/day) than the supplemental dose of LA alone (2.0g/day), the EPO may have caused less competition for the 6-desaturation of α-linolenic acid than that elicited by LA. The finding that the supplementation with LA, but not EPO, reduced the plasma DHA concentration could have important clinical implications. DHA has been suggested to prevent or retard the development of atherosclerotic cardiovascular diseases [24] and considerable evidence indicates that patients undergoing hemodialysis have a high incidence of coronary artery disease and accelerated atherosclerosis [25]. To our knowledge, no previous animal or human studies have addressed the effects of supplementation with EPO or LA on plasma concentrations of EFAs in patients with ESRD. Although our data were obtained from a small number of patients, supplementation with EPO seems safe with regard to the influence on the plasma concentration of DHA.

We demonstrated that oral supplementation with EPO at the dose employed herein (i.e., 2 g/day) significantly (p < 0.05) improved skin scores during the 6-week intervention period, whereas LA did not (Fig. 2). In addition, those given EPO showed a trend toward a greater improvement than those given LA (Fig. 2) in pruritus scores. Recently, Peck et al. [26] reported that oral supplementation with fish oil at a dose of 6 g/day for 8 weeks increased plasma concentrations of eicosapentaenoic acid (EPA) and DHA and showed a trend toward a greater improvement (0.05 < p < 0.1) in uremic pruritus than those given olive oil or safflower oil. While fish oil is rich in n-3 EFAs (e.g., EPA), olive oil and safflower oil used as the controls are rich in oleic acid (n-9 EFA) and LA (n-6 EFA), respectively. Thus, their results appear to indicate that the oral supplementation with EPA or certain n-3 EFAs may be more effective than those with n-9 or n-6 fatty acids in alleviating uremic skin symptoms. However, because the conversion of LA to its further n-6 EFAs (e.g, GLA, DGLA) is slow (Table 2), it is difficult to say whether the supplementation with fish oil or GLA-rich EPO would have more salutary or favorable effects on uremic skin symptoms. Nonetheless, these two studies clearly showed that oral supplementation with oil products being rich in n-3 (e.g., fish oil) or n-6 (e.g., EPO) EFAs may correct or alter the plasma concentrations of EFAs in patients undergoing hemodialysis. Clinical trials for assessing certain therapeutic modality for the treatment of uremic skin symptoms are difficult to carry out, because of the inherent variability of the clinical state, the subjective nature of assessment and a large placebo response. Thus, the studies of Peck et al. [26] and ours warrant further studies to assess the efficacy and safety of oral supplementation with EFA-rich oil products as an alternative therapeutic option for the treatment of uremic skin symptoms.

The present study may shed some light on the pathogenesis of uremic skin symptoms based upon the biochemical and clinical responses to supplementation with EPO. The favorable effects on uremic skin symptoms produced by the administration of EPO containing 0.18 g/day of GLA (Fig. 2) was coincident with increased plasma concentrations of GLA and DGLA. Interestingly, 7 previous studies [19, for review see ref. 27], with one exception [28], showed that the oral administration of EPO (ca. 0.4 g/day of GLA) produced a greater improvement in one or more symptoms or signs of atopic dermatitis than placebo. Clinical symptoms of atopic eczema (e.g., dryness of the skin and pruritus) are similar, if not identical, to uremic skin symptoms. Patients with atopic eczema were shown to have elevated and reduced plasma concentrations of LA and DGLA, respectively, indicating that the conversion of LA to its further n-6 EFA metabolites (i.e., Δ-6-desaturation) would be impaired [19,22]. In contrast, dialysis patients were shown to have reduced plasma concentrations of LA and AA [10,11]. Based upon these findings we are tempted to speculate that the two different groups of patients might possess a reduced plasma concentration of DGLA as a common abnormality in plasma n-6 EFA. LA is converted to GLA by Δ-6 desaturase and then to DGLA by elongase (Fig. 1). DGLA is metabolized not only to an anti-inflammatory eicosanoid (i.e., prostaglandin E1) via cyclooxygenase [13] but also to 15-OH-DGLA which possesses potent inhibitory activity against 5- and 12-lipoxygenases that produce pro-inflammatory leukotrienes from AA [15] (Fig.1). Interestingly, the oral supplementation of EPO elicited no appreciable changes in plasma AA concentrations, thereby increased plasma DGLA/AA ratio of 0.2 observed at baseline to 0.23 and 0.25 at 6 and 12 weeks after the EPO therapy began, respectively (Table 2). Assuming that inflammation in cutaneous tissue plays an important role in the development or aggravation of pruritus and erythema not only in patients with atopic eczema but also in dialysis patients, EPO-induced elevation in the plasma concentrations of GLA and DGLA (Table 2) may not only augment the synthesis of an anti-inflammatory eicosanoid (i.e., PGE1) but also suppress the synthesis of pro-inflammatory eicosanoids, thereby ameliorating some of the uremic skin symptoms (Fig. 2) in these patients. In support of this hypothesis previous reports demonstrated that the administration of GLA to human volunteers and patients with rheumatoid arthritis increased the DGLA levels in plasma, mononuclear cells and platelets as well as enhanced the PGE1 production from stimulated monocytes and reduced prostaglandin E2 and leukotriene B4 synthesis [29,30]. In addition, the administration of GLA produced a significant improvement in their inflammatory symptoms [31,32]. However, our explanation for a favorable effect of EPO on uremic skin symptoms is at present hypothetical, because we did not measure plasma or tissue concentrations of the above-mentioned eicosanoids in our patients. Further studies are required to clarify the linkage between changes in the plasma DGLA and tissue levels of pro- and anti-inflammatory eicosanoids in patients with ESRD suffering from uremic dermatoses. In addition, because we did not measure plasma DGLA levels in healthy subjects in the present study, we cannot answer the questions as to whether our hemodialysis patients have a depleted plasma DGLA and whether such a change per se would have an etiological implication for uremic skin symptoms.

Regarding the effects of supplementation with EPO or LA on dryness of the skin, changes in skin scores obtained from the two treatments were apparently similar (Fig. 2). However, the statistical analysis performed on the changes observed over the study period showed that the EPO treatment produced a statistically significant (p < 0.05) improvement but the LA treatment did not (p = 0.08). In addition, the administration of LA gave rise to a significant elevation in plasma concentration of LA, but that of EPO did not (Table 2). These data seem contradictory to the idea that a deficiency of LA in the cutaneous tissue per se may be associated with dryness of the skin in patients undergoing hemodialysis.

In conclusion, we observed that oral supplementation with EPO containing 0.18 g/day of GLA over 6 weeks gave rise not only to a significant elevation in the plasma concentration of DGLA but also to a trend toward a greater improvement than that of LA in uremic skin symptoms. However, it remains unclear whether changes that occurred in the concentrations of plasma EFAs and in the clinical responses are related each other by certain unidentified biochemical mechanism(s). Nonetheless, we consider that the present study warrants further studies with greater numbers of patients to confirm our promising preliminary results.

Acknowledgments
The authors thank nurses in Toyama Clinic who helped us for completing the present study.

References

1. Rosen T: Uremic pruritus: a review, Cutis1979;23:790-792.
2. Bencini PL, Montagnino G, Citterio A, Graziani G, Crosti C, Ponticelli C: Cutaneous abnormalities in uremic patients. Nephron 1985;40:316-320.
3. Ponticelli C, Bencini PL: Uremic pruritus: a review. Nephron 1992;60:1-5.
4. Nielsen T, Andersen KE, Kristiansen J: Pruritus and xerosis in patients with chronic renal failure. Dan Med Bull 1980;27:269-271.
5. Hampers CL, Katz AI, Wilson RE, Merrill JP: Disappearance of 'uremic' itching after subtotal parathyroidectomy.N Engl J Med 1968;279:695-697.
6. Matsumoto M, Ichikawa K, Horie A: Pruritus and mast cell proliferation of the skin in end stage renal failure. Clin Nephrol 1985;23:285-288.
7. Francos GC, Kauh YC, Gittlen SD, Shulman ES, Besarab A, Goyal S, Burke JF Jr: Elevated plasma histamine in chronic uremia. Int J Dermatol 1991;30:884-889.
8. Hartop PJ, Prottey C: Changes in trans-epidermal water loss and the composition of epidermal lecithin after applications of pure fatty acid triglycerides to the skin of essential fatty acid deficient rats. Br J Dermatol 1976;95:255-264.
9. Wene JD, Connor WE, den Besten L: Development of essential fatty acid deficiency in healthy men fed fat-free diets intravenously and orally. J Clin Invest 1975;56:127-134.
10. Ahmad S, Dasgupta A, Kenny M: Fatty acid abnormalities in hemodialysis patients: effects of L-carnitine administration. Kidney Int 1989;36 (suppl 27):S-243-S246,
11. Dasguputa A, Kenny MA, Ahmad S: Abnormal fatty acid profiles in chronic hemodialysis patients: possible defficiency of essential fatty acids. Clin Physiol Biochem 1990;8:238-243.
12. Bowser PA, White RJ: Isolation, barrier properties and lipid analysis of the stratum compactum, a discrete layer of the stratum corneum. Br J Dermatol 1985;112:1-14.
13. Lovell CR, Burton PA, Duncan EH, Burton JL: Prostaglandins and pruritus. Br J Dermatol 1976;94:273-275.
14. Fantone JC, Kunkel SL, Zurier RB: Effects of prostaglandins on in vivo immune and inflammatory reactions; in Goodwin JS (ed): Prostaglandins and Immunity, Boston, Martins Nijhoff Publishing, 1985, pp 123-146
15. Miller CC, Ziboh VA, Jones AD: Guinea pig epidermis synthesis 15-hydroxy-8,11,13-eicosatrienoic acid (15-OH 20:3n-6) from dihomogammallinolenic acid: a potent lipoxygenase inhibitor derived from dietary primrose oil. J Invest Dermatol 1987;88:507.
16. Brenner RR: Nutritional and hormonal factors influencing desaturation of essential fatty acids. Prog Lipid Res 1982;20:41-48.
17. Daugirdas JT, Depner TA: A nomogram approach to hemodialysis urea modeling. Am J Kidney Dis 1994;23:33-40.
18. Lowrier EG, Lazarus JM, Hampers CL, Mrrill JP: Some statistical methods for use in assessing the adequacy of hemodialysis. Kidney Int 1975;2 (suppl):S231-242.
19. Shalin-Karrila M, Mattila L, Jansen CT, Uotila P: Evening primrose oil in the treatment of atopic eczema: effect on clinical status, plasma phospholipid fatty acids and circulating blood prostaglandins. Br J Dermatol 1987;117:11-19.
20. Manku MS, Morse-Fisher N, Horrobin DF: Changes in human plasma essential fatty acid levels as a result of administration of linoleic acid and gamma-linolenic acid. Eur J Clin Nutr 1988;42:55-60.
21. Lasserre M, Mendy F, Spielman D: Effects of different dietary intakes of essential fatty acids on 20:3n-6 and 20:4n-6 serum levels of human adults. Lipids 1985;20:227-233.
22. Manku MS, Horrobin DF, Morse N, Wright S, Burton JL: Reduced levels of prostaglandin precursors in the blood of atopic patients: defective delta-6 desaturase function as a biochemical basis for atopy. Prostaglandins Leukotriens Med 1982;9:1-13.
23. Horrobin DF, Manku MS: Clinical biochemistry of essential fatty acids. in Horrobin DF (ed): Omega-6-essential fatty acids: Pathophysiology and roles in clinical medicine. New York, Alan R.Liss, Inc., 1990, pp 21-54
24. Burr ML, Gilbert JF, Holliday, RM, Elwood PC, Fehily AM, Rogers S, Sweetnam PM and Deadman NM: Effects of changes of fat, fish, and fiber intakes on death and myocardial reinfaction: diet and reinfarction trial (DART). Lancet 1989;2:757-761.
25. Rostand SG, Gretes JC, Kirk KA, Rutsky EA, Andreoli TE: Ischemic heart disease in patients with uremia undergoing maintenance hemodialysis. Kidney Int 1979;16:600-611.
26. Peck LW, Monsen ER, Ahmad S: Effect of three sources of long -chain fatty acids on the plasma fatty acid profile, plasma prostaglandin E₂ concentrations, and pruritus symptoms in hemodilaysis patients. Am J Clin Nutr 1996;64:210-214.
27. Wright S: Essential fatty acids and the skin. Brit J Dermatol, 1991;125:503-515.
28. Bamford JTM, Reiner RW, Gibson CM: Atopic eczema unresponsive to evening primrose oil. J Am Acad Dermatol, 1985;13:959-963.
29. Callegari P, Zurier RB: Botanical lipids: potential role in modulation of immunologic resposnses and inflammatory reactions. Rheum Dis Clin North Am 1991;17:415-425.
30. Pullman-Mooar S, Laposata M, Lem D, Holman RT, Levental IJ, DeMarco D, Zurier RB: Alteration of the cellular fatty acid profile and the production of eicosanoids in human monocytes by gamma-linolenic acid. Arthritis Rheum 1990;33:1526-1533.
31. Tate G, Mandell BF, Laposata M, Ohliger D, Baker DG, Schumacher HR et al.: Suppression of acute and chronic inflammation by dietary gamma linoenic acid. J Rheumatol 1989;16:729-734.
32. Leventhal LJ, Boyce EG, Zurier RB: Treatment of rheumatoid arthritis with gammalinolenic acid. Ann Intern Med 1993;119:867-873.

Figure 1
Metabolic pathways of omega-3 (n-3) and n-6 fatty acids and their relevant metabolites possessing pro- and anti-inflammatory properties. The pathways consist of progressive desaturation at delta-5 and -6 positions alternating with elongation (addition of two carbons). Abbreviations: LTB₄ = leukotriene B₄, PGE₁ = prostaglandin E₁, PGE₂ = prostaglandin E₂, 15-OH DGLA = 15-hydroxydihomoγ-linolenic acid