[page 85↓]

4. Discussion

The achievement of this research has been the selection of appropriate methods for the specific requirements of the Middle East. Many others have worked on the improvement of PCR-diagnosis of leishmaniasis in other regions of the world, and much has been achieved in the past decade. Based on the experiences made within this research, different aspects of extraction- and PCR-methods will be discussed, which can be used as a basis for future applications.

The PCR methods are discussed chronologically, as they have been gradually established in the laboratory. The results are mainly discussed within the sections of the PCR-methods by which they were obtained. This proceeding was favoured because it reflects the actual situation as it was, where neither the methods nor the samples were available at the beginning.

The collection of samples depended greatly on the contacts made during the study. Sometimes coincidences led to the collection of new sample groups, as was the case with the new L.tropica focus in the West Bank (Wadi Albethan): A dermatologist in Nablus had invited us to visit. The same day he had treated a patient with CL from Wadi Albethan. Through the initiative and help of Samir Sawalha (Masters student of the Al Najar University, Nablus), the first visit to families in the village was arranged immediately and more visits followed.

4.1. Sampling and preservation:

4.1.1. Collection of dermal scrapings:

The detection rate of leishmanial DNA by PCR in dermal scrapings proved to be higher than in microscopical smears, as especially seen in two study groups (desert rodents and patients from Wadi Albethan) The detection rate of PCR exceeded the detection rate of microscopy on dermal scrapings and skin biopsies also in previous studies (Belli et al ., 1998; 1999; Rodriguez et al. , 1994). It is important that the sampling be performed correctly, according to the instructions given by Herwaldt, (1999) (chapter 2.2.1.). Belli et al ., (1998) found the PCR on dermal scrapings (Chelex-extracted) to be 100% sensitive and also to 100% specific. Since dermal scrapings are less invasive and much easier to collect, he suggests that they can fully replace skin biopsies. This was also maintained by Rodgers et al ., (1990), who suggested that patients should be spared from skin biopsies, especially if lesions are located in the face. Instead of sterile surgical blades, sterile wooden toothpicks can be used as well for sampling (Belli et al ., 1998; Harris et al ., 1998). It is known that skin biopsies have a higher detection rate than dermal scrapings if examined microscopically (Andresen et al ., 1996), but with PCR the sensitivity is high enough using dermal scrapings.


[page 86↓]

4.1.2. Specimens other than dermal scrapings for direct PCR-diagnosis:

Apart from dermal scrapings, peripheral blood has been successfully used for diagnosis in this study. Others have previously shown that many different kinds of specimens can be used for PCR-diagnosis of leishmaniasis. Extracts from tissue samples, smears, lymph node and bone marrow aspirates and peripheral blood have been used successfully (Smyth et al. 1992; Lopez et al. , 1993; Eresh et al., 1994; Wilson et al., 1995; Laskay et al., 1995; Bhattacharyya et al., 1996; Osman et al ., 1997; Belli et al., 1998; Noyes et al. 1998). PCR has also been performed on sandflies (Rodriguez et al ., 1999; Aransay et al ., 2000).

4.1.3. Preservation on filter paper:

Blotting of tissue scrapings on filter paper is an excellent method for long-term preservation. The preservation on filter paper proved to be especially suitable for field studies due to very good conservation of the DNA, without cooling, with little storage space and low expenses. No additives for the preservation are required, which could potentially interfere with the PCR. The samples can be delivered to the laboratory weeks after collection without any loss of quality. Filter paper has been used previously for the preservation of peripheral blood spots (Osman et al ., 1997; Campino et al ., 2000), dermal scrapings (Harris et al ., 1998; Färnert et al ., 1999) and different kinds of aspirates (El Tai et al ., 2000).

It has been shown in this study that the filter paper samples were stable for more than one year. The rodent samples had been stored at room temperature for more than one year before they were processed. The amplifications have been very efficient and no loss of quality was observed. A comparative study on Leishmania positive tissue on filter paper, which had been stored at different temperatures (-20 ° C, 4 ° C, 20 ° C) for more than one year, has shown that the DNA was conserved equally well at all temperatures. It can be assumed that humidity affects the stability of the DNA, but this has not been examined here. Freezing of filter paper samples may be preferred in humid climates. (Campino et al ., 2000). Färnert et al. , (1999) have found that blood samples stored at 4 ° and at 20 ° C lost some sensitivity after a period of 30 days, whereas blood samples on filter paper were stable for several months at 4 ° and at 20 ° C. For the storage of blood samples the choice of anticoagulant is relevant. Färnert et al. , (1999) found inhibition in samples extracted from heparin-blood and equally efficient amplification from EDTA- and citrate-blood. Heparin is known to inhibit PCR. EDTA is also known to inhibit PCR, but the amount used for anticoagulation is small enough not to interfere with the reaction. By using filter paper also for peripheral blood (diagnosis of VL), the problem of PCR inhibition by anticoagulatory agents can be avoided.


[page 87↓]

4.1.4. Classical preservation methods:

It was demonstrated that samples differently preserved can be processed as well: Paraffin embedded biopsies (after deparaffinization, see passage 2.2.4.2.) and Giemsa-stained smears were extracted with the phenol-chloroform extraction and successfully amplified. An eventual loss of sensitivity can not be excluded but was not further examined. The formalin-fixed jackal spleens were all negative, either due to true negativity, to DNA degradation prior to the extraction or to inhibition. The biopsied material had not been preserved well, judging from the outer appearance. Formalin-fixed biopsies have been used previously for the diagnosis of leishmaniasis with good success by Mimori et al ., (1998). The positive results with paraffin embedded biopsies and Giemsa-stained smears in this study show that samples other than dermal skin scrapings may also be proceeded (if received from other hospitals). This shows that the same methods can be employed for retrospective studies (Laskay et al ., 1995).

4.2. Discussion of DNA extraction methods:

Several different extraction methods have been employed, ranging from thorough purification to very simple and crude methods. All extraction methods were successful, each of them having advantages and disadvantages. Two objectives had to be pursued, one being the best possible purification with elimination of inhibitory factors, the other being the highest possible yield of DNA. A good balance had to be found between the two requirements. Thorough purification was achieved only through many single steps, always implying a loss of template to a certain extent and an increased risk of sample cross-contamination. Crude extractions did not eliminate inhibiting factors sufficiently (additives can circumvent the problem). Besides, it was important to develop a method, which would readily be accepted in terms of practicability and cost effectiveness (time and material) in a clinical laboratory. Avoidance of hazardous chemicals is another criteria, which might be considered for routine diagnosis.

4.2.1. Phenol-chloroform-extraction:

The phenol-chloroform extraction is a very efficient and thorough purification method, which is comparatively hazardous and time consuming. Many steps of lysis, digestion, extraction and washing require much opening and closing of the 1.5 ml tubes, implying an increased risk of sample cross-contamination. The phenol-chloroform extraction is highly efficient in removing inhibiting factors, such as hemoglobin. It therefore proved to be reliable also for samples with a higher blood content. All samples that were conserved differently (formalin-fixed, paraffin-embedded biopsies and Giemsa stained smears) were proceeded with this method because it had the highest potential to remove inhibitory substances. Apart from that, phenol inactivates other [page 88↓] infective agents (eg. HIV), which may be a major consideration in some regions (eg. south-western Europe). Proteinase-K digestion with subsequent phenol-chloroform extraction is the best known classical extraction method, which was successfully employed also in many previous studies on clinical samples (Smyth et al ., 1992; Ashford et al ., 1995; Andresen et al ., 1996; Osman et al ., 1997; Mimori et al ., 1998; Färnert et al ., 1999; Campino et al ., 2000).

4.2.2. Guanidine-extraction:

The guanidine extraction was efficient and reliable as well. It is especially efficient in the recovery of traces of DNA. The guanidine extraction proved to be less reliable on samples with a high blood content. The procedure is shorter and less steps are required than with the phenol-chloroform-extraction. The centrifugation steps are short (seconds) and a micro-centrifuge suffices. An advantage is also that the reagents are less hazardous than the ones used for the phenol-chloroform extraction. Besides, the samples are only once transferred to fresh 1.5 ml tubes, in contrast to 4 sets of fresh tubes needed during the phenol-chloroform extraction (contamination risk!). The guanidine extraction is a classical method in molecular biology, especially for forensic applications. Only few groups have employed this method in the field of leishmaniasis, also with good results (Noyes et al ., 1998). An interesting study has been conducted by Lachaud et al ., (2001) who compared the phenol-chloroform extraction with the guanidine extraction. In contradiction they found that the guanidine method was slightly more sensitive when whole blood (!) was used, whereas the phenol-chloroform method was superior when buffy coats were extracted. The difference was mainly that the phenol-chloroform extracted buffy coats produced stronger signals, the sensitivity was practically the same.

4.2.3. Chelex-extraction:

The Chelex-extraction was by far the shortest procedure. After over night digestion, heating for 10 minutes and a short centrifugation the supernatant was directly submitted to PCR. Results were obtained already in the early afternoon of the following day. Using the more thorough extraction methods, PCR-results could only be obtained by the evening of the day after the over night incubation. Furthermore, it was shown that an incubation of only 1 hour is sufficient for successful amplification, but the yield might not be maximal. The fact that the samples stay in the same Chelex-suspension from the beginning to the end of the procedure, without opening and closing of the tubes, reduces the risk of sample cross-contamination. Besides, no material is lost during the procedure as it is the case with the phenol-chloroform extraction, in which a small part of the original sample is naturally lost in every transfer of the upper phase. Other experiments with crude extractions have clearly indicated that hemoglobin residues may inhibit [page 89↓] the PCR. A relatively high tissue content and a low blood content (= sample quality with highest success rate) of the skin scrapings reduces the risk of inhibition. The results on the Psammomys ear scrapings have proved the high efficiency of this method, as it has been found as well in previous studies on human samples (Harris et. al ., 1998; Belli et. al ., 1998; 1999). An extensive study on dermal scrapings in 232 patients with American CL has shown that the Chelex extraction was suitable for reliable PCR diagnosis. Using primers 13A/13B, 205 patients were PCR-positive (85%) (Belli et al ., 1999). Especially for larger studies, the Chelex extraction is a very convenient and time saving method. The superior qualities of the Chelex method are: low risk of contamination, saving of labour and material, combined with good sensitivity. For routine use in a clinical laboratory, the employment of the Chelex method could reduce the costs for the PCR-diagnosis of leishmaniasis enormeously. Several hours of bench work can be avoided and laboratory supplies can be spared (pipette tips, 1.5 ml tubes and chemicals). The Chelex method is limited in its capacity if inhibitory substances are present (has to be evaluated yet). For the diagnosis of VL it can be suggested to use the buffy coat (also spotted on filterpaper).

4.2.4. Crude extraction methods:

It has been shown that proteinase-K digestion and boiling (5 minutes) in a water bath was sufficient for successful PCR-amplification from dermal scrapings. These experiments were performed in the beginning of the study, when primers Uni21/Lmj4 were tested for their sensitivity on clinical samples. The fact that results were obtained with the least sensitive primers employed in this study is promising for diagnosis with the more sensitive primers. After the Chelex extraction was found to work well, experiments on other crude preparations became obsolete for this study. Crude DNA-preparations have been tested previously by others, also with good success (Rodriguez et. al ., 1994; Laskay et al. , 1995; Harris et al ., 1998; Belli et al . 1998). There are more variations and combinations, which would be worth-while to test, as for example proteinase-K digestion in 5% Chelex. Lopez et al ., (1993) have used DNAse I digestion prior to PCR in order to expose more kDNA minicircles of the catenated network. The sensitivity was clearly increased when compared to non-digested samples (not tested in this study).

4.2.5. Extraction methods related to different types of specimens:

It was repeatedly observed that the quality of the results appeared to depend on the extraction method chosen for a certain type of sample. In a cooperative study (Mattheson and Anders, submitted 2001) a series of 7 different specimens were submitted to several different extraction methods in order to find the optimum extraction method for each different type of sample. Primers specific for the cytochrome b gene of the mitochondrial DNA were employed to detect [page 90↓]mammalian DNA. As expected, different extraction methods performed differently, depending on the types of specimens. The phenol-chloroform extraction was superior with the blood containing samples (1 drop of whole blood and of EDTA blood on filter paper, 500 µl of whole blood). The guanidine extraction was more efficient than the phenol-chloroform extraction with macerated raw tissue. The chelex method was almost as efficient as the phenol-chloroform extraction with blood spots on filter paper but it was much less efficient with larger amounts of blood and raw tissue. This comparative study showed that it is worth-while to select the extraction method according to the amount and type of sample.

4.2.6. Storage of extracted samples:

The extracted DNA was usually resuspended in ddH 2 O. The samples were directly submitted to PCR for best results. At 4 ° C the DNA degraded within days. Therefore aliquots of the extracted DNA were stored at -20 ° C, in case additional reactions were necessary. Usually good results were obtained also after one round of freezing and thawing. It was considered to use TE buffer (pH 8) instead of ddH 2 O, in order to stabilize the DNA. It was not used here because only small amounts of template would have been tolerated in the PCR-reaction due to the inhibitory activity of the EDTA present in the buffer. Since the phenol-chloroform extracted samples were resuspended in a much smaller volume than the guanidine- and Chelex-extracted samples, TE could be used for the phenol-chloroform extracted samples without any problems (advantage). The DNA is evidently more concentrated and less amounts of template are required for successful amplification, the inhibitory activity of EDTA is negligible then. When a small amount of TE-buffer (eg. 5 m l) was added to extracted samples the PCR was not inhibited but the possibly higher stability of the DNA has not been examined yet.

4.3. Reliability of the employed PCR-methods:

4.3.1. Sensitivity:

The kinetoplast specific primers have been selected because of their potential for high sensitivity. The aim of the direct PCR diagnosis was the detection of single parasites from clinical material. Studies on purified DNA of reference strains have confirmed the high sensitivity of primers 13A/13B and MPL1/MP3H up to less than single parasites (Rodgers et al ., 1990; Lopez et al ., 1993). Experiments with serial dilutions of DNA within this study have found sensitivities with primers 13A/13B up to 10 fg, primers Uni21/Lmj4 up to 1 pg (25 fg DNA with Gels Star staining) and up to 1 pg with primers LITSRn/L5.8S (ITS-1). Even though primers Uni21/Lmj4 were theoretically capable of detecting single parasites it was often not seen. When clinical samples are examined less sensitivity is expected due to inhibitory agents. [page 91↓] Rodgers et al ., (1990) found the sensitivity to be 10-100 fold less with clinical samples, compared to purified DNA from cultured strains (inhibition). Similar observations have been made repeatedly during this study. This fact could not be evaluated in a quantitative way since the intracellular amastigotes of a clinical sample would have to be compared with defined amounts of purified DNA which is practically impossible. A comparison between purified DNA and a defined number of promastigotes in peripheral blood is certainly the best possible comparative approach as recently performed by Dinse et al ., (2001). The factor of cell lysis necessary to release the intracellular amastigotes in a real sample is not considered. Furthermore, dermal scrapings can not be standardized, every new sample has a different consistency (eg. amount of hemoglobin), which greatly influences the outcome. In order to assess the quality of the PCR in a more realistic way, results were compared with the results of the corresponding smears. The study with primers 13A/13B on desert rodents (Chelex-extracted) has clearly demonstrated a sensitivity which is unlikely to be surpassed. Primers MPL1/MP3H amplified even a smaller part of the kinetoplast minicircles (70 bp) and were therefore expected to have a similar sensitivity on clinical samples as primers 13A/13B. This was confirmed in the 6 patients who had contracted CL in Tuiji (Bolivia). The ITS-1 primers LITSRn/L5.8S were found to be highly sensitive on clinical samples since they were able to amplify many samples, which had been positive by primers 13A/13B and/or MPL1/MP3H earlier. This can be stated since the samples had been extracted earlier and had to be defreezed for this PCR. Freezing may lead to deterioration of the DNA quality.

In other studies, the sensitivity of PCR was always superior to the sensitivity of microscopy (Belli et al ., 1998; Andresen et al ., 1996; Rodriguez et al ., 1994). PCR was also found to be more sensitive (100%) than serology (63%) (Ashford et al ., 1997). It has been repeatedly shown that the potential sensitivity of PCR reached the detection of single parasites (Rodgers et al ., 1990; Lopez et al ., 1993; Noyes et al ., 1998; Harris et al ., 1998).

4.3.2. False negatives/ inhibition:

It was observed repeatedly that an increase of template (DNA, lysate or supernatant) in the reaction mix did not necessarily lead to a better result. On the contrary, it became obvious that beyond a certain amount of template DNA the amplification was inhibited. This applied also for well purified samples. In the first attempt to apply the direct PCR diagnosis on the 23 patient samples from Wadi Albethan 10 m l of template (10 m l out of 100 m l extracted, resuspended DNA) were used resulting in only 12 positives. In the repeat PCR (performed by Kefaya Azmi) only 2 m l were used resulting in 23 positives (100%). Several experiments showed that the inhibition can be controlled by adding either BSA (eg. 5 m l of a 10mg/ml solution) or DMSO [page 92↓] (2.5%) together with formamid (1%) to the reaction. Using these additives, the PCR tolerates more µl of template DNA, which is of interest especially for the less sensitive PCR (Uni21/Lmj4). Inhibition especially with crude samples was observed by others as well. Belli et al ., (1998) circumvented the problem by diluting the samples up to 10-fold. Since the employed primers (13A/13B and MP3H/MPL1) were highly sensitive, the template in the diluted samples sufficed. With sensitive primers a small amount of template DNA (1-2 m l) is sufficient to ensure amplification.

The volume of the resuspension is also a factor worth to consider: A smaller volume is of advantage because the extracted DNA is more concentrated and therefore more stable. Only with the phenol-chloroform extraction could the volume be reduced to 25 m l. With the other methods the volume could not be reduced for technical reasons. The guanidine extracted samples were resuspended in a higher volume (100 m l) to allow the silica beads to be completely in suspension, in order to release the attached DNA into the solution. The Chelex extraction required at least 200 m l of 5% Chelex, otherwise the filter paper would not have been covered completely during the incubation.

False negative samples could be also due to uneven distribution of Leishmania DNA in an extracted patient sample, eg. if a sample contains only 1-10 parasites not every single m l of the resuspended DNA contains necessarily enough template to support amplification. It can be considered to amplify different amounts of one sample at the same time to avoid false negativity. Another possibility is to include an inhibition control, using primers which target human DNA. Human DNA is present in excess in all clinical samples and PCR-amplification can thus be controlled.

4.3.3. False positives/ contamination:

The main problem of sensitive PCR methods is the risk of contamination, leading to false positive results. Contamination was monitored by including at least two negative controls in every extraction series. A piece of plain filter paper was used also as a control of the filter paper itself. A human blood sample on filter paper was used to rule out any non-specific amplification of human DNA. For the Psammomys samples ear tissue was used from a negative laboratory animal. The amplification was controlled for any contamination related to the PCR itself (reaction mix without template).

It appeared that contamination occurred mostly during the extraction, with negative extraction controls being positive and negative PCR controls being negative. For this reason several series of extractions (Hadassah patients) had to be excluded from the study. The risk of sample cross-contamination was minimized, by following a number of precautions: Bleaching and UV- [page 93↓] radiating of surfaces and pipettes were the most important measures to avoid contamination. After these measures were followed strictly, the problem was controlled and extraction controls as well as PCR controls were then negative. The risk of contamination is known to be the main problem of PCR, which has been discussed also by others in relation to PCR-diagnosis of leishmaniasis (Adhya et al ., 1995; Wilson , 1995; Belli et al ., 1998).

4.3.4. Reproducibility of PCR-results:

In repeated extraction and PCR (13A/13B) from single drops of the same blood samples (4 dogs with CVL) the same intensities of bands were reproduced. This concordance suggested a high correlation between the parasitic load and the amount of amplified product. A positive correlation between the intensity of bands and the number of parasites seen by microscopy was also found by Andresen et al ., (1996). The distribution of amastigotes within a lesion is not homogenous. One experiment with three dermal scrapings which had been collected as a series from three different areas of the same lesion, showed great differences in the intensities of amplified bands. It is therefore also possible that the smear contains amastigotes and the filter paper not, or reversely. Also the relative amount of blood influences the results. Bloody samples often contain less parasites, since the blood does not necessarily originate from the lesion itself. If parasites are equally distributed in the sample, as it is the case in peripheral blood of VL patients the PCR seems to be reproducible to a high extent. .

4.4. Discussion of PCR results:

4.4.1. Results with primers Uni21/Lmj4:

The PCR using the kDNA primers Uni 21/Lmj4 was found to be useful to distinguish between L.major and L.tropica infections. The PCR on purified DNA and on cultured organisms proved to be reliable, yielding always a 680 bp product in L.major typed strains and a product of 820 bp in L.tropica typed strains. Non-leishmanial Kinetoplastidae did not amplify. The sensitivity with purified DNA (from cltured promastigotes) was promising, with regard to the potential use for direct diagnosis. Harris et al. (1998) estimated that 1 pg of DNA is equal to 10 parasites. The potential sensitivity of primers Uni21/Lmj4 therefore was corresponding to the DNA of 1/4 parasite. Nevertheless, the primer pair Uni21/ Lmj4 was often not sensitive enough to detect single parasites from clinical samples. Efforts have been made to optimize the sensitivity by making the following changes:

1. DMSO and formamid or BSA were added to the PCR in order to block inhibitory activity. Thus more µl of extracted template were tolerated in the reaction.


[page 94↓]

2. The choice of the polymerase influenced the outcome to a great extent: the Promega Taq polymerase often failed to amplify the 680-850 bp sequence from clinical samples. The TaKaRa Ex Taq and the Fermentas Taq polymerase proved to be more reliable.

3. It was hoped the sensitivity would be improved under less stringent PCR-conditions. A lower annealing temperature (50 ° C instead of 60 ° C) yielded also good results. Since occasionally non-specific PCR products of different sizes appeared the original annealing temperature (60 ° C) was finally adapted as the standard annealing temperature.

4. A higher number of cycles (45 instead of 35) was an alternative option to increase the sensitivity.

5. The staining with Gel Star stain instead of ethidium bromide staining clearly increased the sensitivity. Bands which were nearly invisible with ethidium bromid were distinct and clear when stained with Gel star stain.

The PCR with primers Uni21/Lmj4 showed primarily good results with dermal scrapings on filter paper, and appeared to be a suitable method for direct diagnosis and simultaneous differentiation of L.major and L.tropica . Even on crude extractions (from dermal scrapings on filter paper), samples were amplified successfully (lysis with NET-buffer and proteinase K digestion). At the beginning, the results were very promising, but later this PCR often failed to produce results, even on microscopically confirmed patients. Nevertheless, whenever the PCR succeeded the results were valuable and very helpful. The results were collected during the developmental stage when conditions of extraction and PCR were often modified. No amplification was probably often due to sub-optimal conditions (too much template eg.). A statistical analysis was therefore not possible. In 6 out of 24 PCR-positive patients the diagnosis was made by PCR only, since smear and culture had been negative.

Specific cases diagnosed with kinetoplast primers Uni21/Lmj4:

1.The case of the patient with the severe nose infection and a possible double infection will be discussed in the following section: controversial results.

2.The diagnosis of L.major in the couple from Holon was important to exclude any other causative species. The area is not endemic for leishmaniasis and the source of the infection remained unclear. The couple claimed they had not travelled in- or outside the country since more than one year. The ecology (city, close to the Mediterranean) did not suggest a new focus of L.major . The only conclusive hypothesis was that the neighbour had imported infected sandflies in his car. The couple reported the neighbour was going frequently to the Jordan valley by car. The source of the infection remained unclear.


[page 95↓]

3. The correct identification of L.major from a lesion caused by a L.major vaccination strain confirmed the specificity of the primers (patient No. 24 in Table No. 6).

4. The results from 4 skin scrapings collected from Psammomys ears show that the kDNA PCR with the primer pair Uni21/Lmj4 can be used for detecting infections in reservoir animal species as well.

5. In the case of the infected sandfly from Kfar Adumim L.tropica was confirmed with primers Uni21/Lmj4 as soon as promastigotes were discovered in the medium. This case demonstrated the usefulness of primers Uni21/Lmj4 on newly collected strains. Differentiation between L.major and L.tropica can be achieved long before the normal harvesting of cultures in the stationary phase of growth. Kfar Adumim is a known focus of Leishmania tropica (Klaus et al. , 1994). It is no contradiction that Kfar Adumim residents were found to be infected with L.major as well. Transmission of L.major within the settlement Kfar Adumim has not been proven, but it can also not be excluded. The fact that both, L.major and L.tropica were diagnosed from residents of Kfar Adumim ( L.tropica also from sandfly derived cultures), emphasizes the necessity of species distinction, even within one settlement. The Jordan Valley (Jericho) as the main transmission area of L.major in the country, is very close, the Wadi Kelt being in a distance of less than 1 km. The Bedouins in the area are familiar with CL. The exact limits of L.major transmission are not known in the Judean Desert, where Wadis directly connect with the Jordan Valley.

6. A cultured strain obtained from Costa Rica (via Lionel Schnur, Kuvin Center), yielded a PCR-product of 800 bp. Since the L.braziliensis- as well as the L.mexicana -complex are not amplified with primers Uni21/Lmj4 the only possible New World species amplified with these primers, is L.d.chagasi . The fact that the strain was derived from a patient with VL who is living in an area endemic for L.d.chagasi, corroborated the diagnosis. The same case has been reported as the first case of VL in Costa Rica by Carrillo et al, (1999), L.d.infantum/chagasi has been identified by isoenzyme analysis. This demonstrates that the primers Uni21/Lmj4 may be useful in the New World too. L.d.chagasi infections could be identified and distinguished from all other New World species. This may be of interest especially for the diagnosis of “atypical CL“ in the New World caused by L.d.chagasi (Ponce et al ., 1991).

Controversial results:

In the patient with the severe infection of the nose, the direct PCR diagnosis revealed an infection with L.tropica from one lesion, and possibly a double infection ( L.major/L.tropica ) from the other lesion. The PCR on the culture of the same patient some weeks later revealed a [page 96↓] band of 680 bp ( L.major ). The EF performed on the culture identified this isolate as being L.major . A contamination of the filter paper extraction is presumably not the cause, since all controls were as expected and all other patient samples in the same series produced the smaller fragment of 680 ( L.major ). The more conclusive explanation is a mixed infection with two species. The origin of the patient (Hemdat) in the bordering area of the northern Jordan Valley and the hilly slopes of Samaria (West Bank) is indeed an area, where both species ( L.major and L.tropica ) may occur. Foci of L.major (Jordan Valley) and the newly identified focus of L.tropica in the hilly area (Wadi Albethan) are both very close. Klaus et al. (1994) made similar observations in two patients who had been infected in Kfar Adumim: the PCR (with primers Uni21/Lmj4) had revealed L.tropica from the early cultures. The culture had been re-examined after some time, and L.major had been identified. It has been observed that the initial growth of the culture was very slow, whereas the later culture was thriving well. The most probable explanation was a mixed infection with two species. It is a known fact that some species thrive better than others in culture. Strejan (1963) observed that different species of Leishmania require a different nutritional medium in order to grow well. In a mixed culture one species is therefore graduall y outgrowing the other. Harris et al ., (1998) reported that L.major outgrows L.tropica, as L.mexicana outgrows L.braziliensis in mixed cultures. This implies that diagnostic methods based on cultured parasites (isoenzyme electrophoresis, analysis of the EF, other PCR-methods) might be misleading in the case of double infections. As it was the fact in this patient the classical methods would have missed the more problematic causative agent ( L.tropica ), which requires medical intervention. A concurrent infection with L.major and L.donovani has been reported from Kenya (Mebrahtu et al ., 1991). It can be assumed that double infections were greatly underreported in the past because diagnosis relied mostly on culture-based techniques. The presented case clearly demonstrates the relevance of direct diagnosis.

It can not be excluded that some strains have characteristics of two species: Oren et al. , (1991) reported about a 21 year old Israeli with VL, who came from the area south of the Sea of Galilee. A strain was isolated which was typed as L.tropica by EF whereas isoenzyme analysis had revealed characteristics from both, L.donovani and L.tropica.

The PCR with primers Uni21/Lmj4 can not discriminate between L.tropica and L.d.infantum . A PCR-product of 800 bp was attributed to L.tropica , but theoretically it could be also due to L.d.infantum . There is no record of cutaneous L.d.infantum strains in Israel and the PA so far, but since it has been reported from other countries (Lebanon, Italy, Tunisia, Iran) it has to be considered (Nuwayri-Salti et al ., 1994; Gramiccia et al ., 1989; 1991; Hatam et al ., 1997, respectively). Endemic areas of L.tropica and L.d.infantum are partly overlapping in the Jenin district, which makes the differentiation relevant. An attempt to overcome the limitation in [page 97↓] species discrimination with primers Uni21/Lmj4 was performed by Kefaya Azmi and Obeida Yussuf (students of the Al Quds-University, Jerusalem, laboratory work performed at the Kuvin Center), who tested many restriction enzymes for RFLP- analysis of the PCR product and were able to differentiate the species according to their patterns.

4.4.2. Results with the genus specific kDNA-primers 13A/13B:

The genus specific primer pair 13A/13B proved to be highly sensitive, up to 10 fg (1 pg = 10 parasites, Harris et al ., 1998) and was able to detect probably every single infection. The high sensitivity of the PCR with primers 13A/13B had been also found in previous studies (Rodríguez et al ., 1994). This PCR proved to be an excellent screening method for the presence of Leishmania DNA, directly from dermal scrapings. Noyes et al ., (1998) discussed, that the amplification of the conserved part (120 bp) of the kDNA minicircles is often sufficient because the species is already known (eg. screening of Psammomys ). Primers 13A/13B amplified also non-leishmanial Kinetoplastidae , such as Trypanosoma , which has to be considered if infections are contracted in areas co-endemic with the Chagas disease or sleeping sickness (trypanosomiasis).

The PCR with primers 13A/13B was introduced as a screening method. It has been successfully employed on several sample groups, after the contamination problems were controlled. The specific results will be discussed in extra sections: the results from the Tel Hashomer patients (New World leishmaniasis) will be discussed together with the corresponding results with primers MP3H/MPL1 in the following section, the results from the new L.tropica focus Wadi Albethan (West Bank) will be discussed in section 4.5., the animal samples will be discussed in section 4.6.

4.4.3. Results with L.braziliensis specific primers MP3H/MPL1:

Since New World leishmaniasis is occasionally seen in Israeli travellers it was important to establish also PCR-diagnosis for the Leishmania species of the New World. Because of the risk of mucocutaneous leishmaniasis (MCL, Espundia) later in the case of L.braziliensis -infections, the major goal was to identify the infections caused by the L.braziliensis complex. The diagnosis of L.mexicana species was also desired but of secondary relevance. Primers MP3H/MPL1 were selected because they met the two requirements, being able to distinguish between the L.braziliensis and the L.mexicana complex and also to detect single parasites (Lopez et. al ., 1993). The primers MP3H/MPL1 had been previously employed also on dermal scrapings and had been proved to be more sensitive than microscopy (Belli et. al ., 1998). When tested on reference strains the species of the L.braziliensis complex ( L.braziliensis, L.guyanensis, [page 98↓] L.panamensis ) were amplified very efficiently. L.mexicana and the species of the Old World ( L.major , L.tropica , the L.donovani complex) were not amplified, thus confirming the species-specificity of the primers. The sensitivity of the primers were evaluated empirically on dermal scrapings, in relation to the corresponding results with primers 13A/13B. Anyhow it was important to examine the returners with both primer pairs, since they could have contracted the disease also in Israel, either before or after the journey to America (Table 6, patients 11 and 12!). All 6 patients who had contracted CL in Bolivia were diagnosed with L.braziliensis directly from lesion scrapings. One patient had visited Guatemala and was negative with the L.braziliensis specific primers but positive with the genus specific primers (13A/13B). This finding suggested an infection by species of the L.mexicana complex. L.d.chagasi was another possibility since it would have reacted identically with primers MP3H/MPL1 and primers 13A/13B. So far, “atypical” CL caused by L.d.chagasi has been reported from Costa Rica, Honduras (Ponce et al ., 1991) and from Nicaragua (Harris et al ., 1998; Belli et al., 1999). The other 4 patients who were treated at the Tel Hashomer Hospital had been infected in the Old World. All 4 were negative with the L.braziliensis specific primers and positive with the genus-specific primers, as expected.

Repeated sampling and PCR, before, directly after and again months after treatment showed that the PCR signal became fainter or even disappeared after healing in most cases. In two patients who were suffering from a clinical relapse (patients 1 and 7, Table 8), the previously faint signals became more intense again. This was interpreted as a confirmation of the relapse, requiring another 3 weeks course of systemic pentostam. Not every PCR-positivity months after completion of the treatment should be regarded as a sign of a relapse. Due to the high sensitivity of the PCR a tiny remnant of DNA may easily be amplified, especially if the targeted sequence is as small as 70 bp . A weak amplification was seen in most of the cases after treatment and progressive healing. The clinical appearance needs to be judged primarily and the PCR-results are usually helpful to confirm the clinical findings. One has to be careful to draw conclusions on the variable quantity of the PCR-product. The reproducibility especially from lesion scrapings is limited, as discussed earlier. The amount of collected tissue was not standardized and was different in every new sample. Nevertheless, a constant decrease of PCR-amplification after completion of the treatment is certainly a good indication for cure. A study conducted by (Nuzum et al ., 1995) has in fact shown that 12 out of 13 previously PCR-positive VL patients were PCR-negative 6 months after treatment and cure.

Only recently we learned that a follow up study with repeated dermal scrapings and PCR, as it was performed here, may not be advisable due to increased risk of reactivation. This applies especially for L.braziliensis infections, which tend to recidivate (personal communication with Palmira Guevara, University of Venezuela, Caracas, Venezuela).


[page 99↓]

4.4.4. Results with ITS-1 primers and RFLP-analysis:

The ITS-1 amplification combined with RFLP-analysis was published only recently (Schönian et al ., 2000; El Tai et al ., 2000), otherwise it would have been adopted as the standard methodology of this thesis from the beginning, due to its species specificity and good sensitivity. With purified DNA it had attained the same sensitivity as with primers Uni21/Lmj4 (1 pg). This PCR method allows species specific diagnosis of all Old World species and all New World complexes. The different species of the L.braziliensis complex appear identical, the species of the L.mexicana complex ( L.mexicana and L.amazonensis ) can be differentiated.

In this study the ITS-1-PCR was applied on already extracted clinical samples which had been stored at -20 ° C. The restriction of the PCR-product with the enzyme BsuRI ( Hae III ) showed the expected patterns, allowing for species specific diagnosis. The results of the pre-diagnosed samples were all confirmed. The ITS-1-amplification in combination with RFLP analysis is clearly superior to the other PCR methods presented here. The additional step of about 2 hours incubation and an additional gel electrophoresis run is a negligible disadvantage in relation to the benefits. Most PCR-systems found in the literature achieve either high sensitivity or species specificity. Only a few are sensitive and also species specific for very few species simultaneously. With this method nearly every Leishmania species can be identified with high sensitivity.

4.5. Wadi Albethan, a newly identified focus of L.tropica in the West Bank:

A new focus of L.tropica has been identified in Wadi Albethan, an Arab village east of Nablus. Probably the whole mountain range west of the Jordan Valley is prone to be endemic with L.tropica . The area has not been systematically investigated yet. One soldier reported about several other soldiers being infected with CL in a military camp on the same longitudinal line some kilometers south (personal communication). Kfar Adumim is also on the same longitudinal line as Tiberias in the north (both L.tropica foci). The outbreak of CL in Wadi Albethan was confirmed as being caused by L.tropica with primers Uni21/Lmj4 from the 2 successfully cultured strains , RFLP analysis of the PCR-product (performed by Kefaya Azmi), ITS-1 amplification and RFLP analysis and the PPIP-PCR (performed by Carol Eisenberger, Kuvin Center, Jerusalem). It proved to be very difficult to retrieve sufficient template DNA from dermal scrapings for the amplification with primers Uni21/Lmj4. The ITS-1-PCR had not been published at that time, which would have been the superior method.

The microscopical examination of the 23 smears was sensitive to 39 %. This confirmed the findings of many other studies, in which the microscopical detection rate did not exceed the [page 100↓]sensitivity of PCR. Four smears of this study group, as well as many others seen during the study period, had to be classified as negative, even though the microscopical picture was highly suggestive of a leishmanial infection. Cell like structures in the size of amastigotes were often seen in groups either intracellularly in macrophages or extracellularly. Nuclei and kinetoplasts could not be identified. In most of these cases patients had been treated previously either with intralesional pentostam or liquid nitrogen. It was assumed that the described structures were destroyed amastigotes. The smears showed a picture of successful destruction of the parasite, which is expected under treatment as well as during a natural healing process.

4.6. Discussion of PCR results from animals:

Desert rodents:

The study on 30 ear scrapings of Psammomys was performed in order to approve the reliability of the direct PCR diagnosis. Besides, it was an interesting field study which revealed a high percentage of Leishmania infections in the Psammomys of the Qziot area in the Negev. The fact that all positive smears (20) were also positive by PCR confirmed the specificity of the PCR (no false negatives). The sensitivity by PCR (28) was far higher than by microscopy (20), as expected. The screening of ear tissue from other desert rodents from the Qziot area revealed a Leishmania infection in a rodent species ( Gerbillus dasyarus ), which was not known previously to host the parasite. A new reservoir host species has been identified by PCR. The faint bands of 120 bp indicated the presence of leishmanial DNA. The result was confirmed by one amastigote found in one of the smears. So far, the findings can only prove the infection of this gerbil species; the question if this species is in fact functioning as a reservoir has yet to be elucidated.

Canids:

The peripheral blood from four examined dogs suffering from CVL was PCR-positive with primers 13A/13B. This small study showed that one drop of blood can be sufficient for diagnosis of VL. It can be assumed that VL could be diagnosed equally well also from human blood (discussed later in the section on diagnosis of VL). Thus suitable laboratory tools have been established for the diagnosis of VL too. As mentioned earlier infantile kala azar is occasionally seen in the West Bank and many dogs are infected in Israel (only poor data of dogs in the West Bank) which shows the relevancy.

All jackal ears (of 20 animals) were negative. Most of the western-blot results of the same animals had been either negative or only borderline positive (according to Gad Baneth, Veterinary School, Hebrew University, Rehovot). The examination of the ears had a low chance for positive PCR-results, but it was considered worth-while to try because of the great relevancy [page 101↓] to identify infections in jackals (see Introduction). The jackal spleens in formalin were all negative by PCR. The tissue was not preserved well, which questioned the quality of the extracted DNA anyhow, as mentioned before. Other possible reasons for the negativity could be either due to inhibition of the PCR by remnants of formalin or to the fact that the animals were not infected with Leishmania .

Hyraxes as a reservoir of L.tropica ?:

The examined hyrax tissue (nose, skin and bone tissue of the parched hyrax and the ear biopsy of a living animal) was negative. One hyrax was parched and it was not certain if DNA could be recovered at all. The other hyrax was not exactly from a known endemic area. It was obvious that systematic trapping of hyraxes in a highly endemic area would be necessary in order to detect potential infections. Wadi Albethan would be an excellent study area since it is hyper-endemic for L.tropica and hyraxes are abundant. Due to the recent political developments this study can not be carried out presently but the methodology is available now in the laboratory to conduct these studies in the future.

Hyraxes (Photo 19) were observed close to Kfar Adumim ( L.tropica focus, Photo 13) in the Judean Desert and in Korazim, a settlement on the northern slope of the lake of Galilee, which is one other new L.tropica focus in the country (patient 5, Table 8, Photo 3). Also in Jordan, hyraxes were observed in the three major foci of L.tropica, and were therefore highly suspected to be the reservoir (Saliba et al ., 1997). Previous findings have supported the hypothesis: L.tropica has been identified in a hyrax in Kenya (Sang et al ., 1992). Hyraxes were observed in a new focus of L.tropica in Kenya (Sang et al ., 1994). Hyraxes have been found infected with L.aethiopica in Ethiopia (Ashford et al. , 1973). The cultured parasites had not been characterized at the time of publication. Since L.tropica has been isolated from dogs in Morocco (Dereure, 1991) and was also found in a rat (Aljeboori and Evans, 1980), both animal species have to be considered as being a potential reservoir as well. It has to be clarified if dogs or rodents are only accidental hosts or if they are functioning as reservoir.

4 . 7. Choice of treatment on the basis of PCR-results:

Species- or at least complex-specific identification is essential for the selection of the appropriate treatment. Early differentiation of L.braziliensis and L.mexicana infections can help to avoid unnecessary hospitalization and treatment of patients infected with L.mexicana . If these patients still require treatment, ambulant care would be sufficient and drugs (ketokonazol eg.) could be administered orally (Rodriguez et al ., 1994). Several patients in Israel might be spared from toxic and costly treatment every year. A three weeks course of intravenous pentostam in Israel [page 102↓] sums up to about $ 10,000 for hospitalization and 500 $ for the treatment (20 mg/kg/day pentostam for 20 days) (personal communication with Dr. Eli Schwartz, Tel Hashomer Hospital, Tel Aviv).

Infections with L.mexicana and the L.braziliensis complex were found to respond differently to certain drugs (Navin et al ., 1992). 90 % of patients with L.braziliensis- infections responded well to sodium stibogluconate whereas only 57% of L.mexicana infections showed improvement. Using ketoconazol 89% of patients with L.mexicana infections responded well compared to only 30% of the patients infected with L.braziliensis . The PCR results with primers Uni21/Lmj4 can also give a direction for the choice of the adequate therapy. If L.major is diagnosed (680 bp), usually no treatment is required. If a >800 bp product is found the patient needs to be treated ( L.tropica, L.donovani complex). The ITS-1 amplification combined with restriction is evidently a suitable method for species specific diagnosis and therefore provides the information necessary to treat a patient adequately.

4.8. PCR diagnosis in VL:

This study focused on CL because VL is rare and most cases occur in the West Bank (Photo 14), where the access to doctors and hospitals is complicated. Even if there were single cases in the past two years they were not necessarily referred to our laboratory. The methodology which has been successful for the diagnosis of CL can be applied as well for the diagnosis of VL. As the small study on 4 infected dogs has shown, VL can be diagnosed directly from peripheral blood. It is of great importance that these methods are available in the country, especially for the diagnosis of human (infantile kala azar) in the West Bank but also for epidemiological studies in canids. The methods are suitable for a clinical laboratory, which is equipped with a PCR machine. Osman et al ., (1997 a) evaluated PCR-diagnosis on peripheral blood spots from confirmed and suspected VL patients in Sudan. The PCR was positive in 70% of the confirmed and in 19% of the suspected cases. In the same study, also different clinical specimens were compared for their detection rate by PCR. Lymph node aspirates, bone marrow aspirates and peripheral blood (finger-prick-blood) of microscopically confirmed VL cases in Sudan were collected on filter paper and extracted with phenol-chloroform. The PCR revealed positivity in 100% of the bone marrow aspirates, in 87% of the lymph node aspirates and in 70% of the peripheral blood samples. Even though the detection rate from blood was the least, as expected, it is an excellent method for first line screening in suspected cases of VL. Thus many patients could be spared from the more invasive diagnostic methods, such as bone marrow- and splenic aspirations. Also Campino et al ., (2000) have found a good detection rate with peripheral blood [page 103↓] spots on filter paper. Out of 20 immunocompromised and previously diagnosed VL patients 15 were confirmed by PCR (SSU-rRNA primers).

The use of the buffy coat (Lachaud et al ., 2001) or separated peripheral blood mononuclear cells (PBMC) (Nuzum et al ., 1995 and Belli et al ., 1998) may be even more promising since the parasites are concentrated in these cells. The cells (PBMC) were separated from peripheral blood and only crudely prepared (lysis in a buffer, digestion with proteinase K and boiled for 15 minutes). Belli et al ., (1999) found 7 out of 9 VL patients to be positive by PCR. Adya et al ., (1995) compared the sensitivity of PCR, ELISA and microscopy in Indian patients with symptoms suggestive of VL. Out of 22 patients 17 were positive by PCR, 14 by ELISA and 7 by smears which had been prepared from spleen aspirates.

Since the reservoir and the vector for VL are prevalent in many places in Israel and in the West Bank, CVL is increasingly seen in domestic dogs in Israel, outbreaks in humans could occur as well at any time. It is therefore good to be technically prepared for reliable diagnosis. Serology is available, but antibodies are not necessarily found in the beginning of the disease. Diagnosis from spleen or bone marrow aspirates is also very sensitive, but can be left for the PCR-negative cases (from peripheral blood).

4.9. Guidelines for future applications:

4.9.1 Sampling, preservation and extraction:

As a result of the observations made during the whole study period and the findings of the comparative study (with Carney Mattheson, not included in this thesis), the following proceeding can be recommended for a clinical laboratory: Dermal scrapings on filter paper are very suitable samples which can be highly recommended for the PCR-diagnosis of CL. For the diagnosis of VL, the buffy coat of peripheral blood appears to be the best specimen, which can be spotted onto filter paper as well. Bone marrow aspirations (more sensitive but also more invasive) can be preserved as second line diagnostic measures.

Filter paper samples consisting mostly of tissue and only of a small amount of blood can be extracted with the Chelex method with excellent results. If specimens contain a larger amount of blood (eg. also dermal scrapings on filter paper, which accidently consist mainly of a drop of peripheral blood) or are conserved differently (fixed in formalin, paraffin embedded, Giemsa stained) the phenol-chloroform extraction should be used. The guanidine method can not offer more than the Chelex method and may be preserved for exceptional samples (bone powder, macerated tissue). It may be considered for samples in which a certain degree of purification is necessary and a loss of material (as it can not be avoided with the phenol-chloroform extraction) would be critical.


[page 104↓]

The problem of inhibition has to be considered. Inhibition is influenced by the original sample size in relation to the volume of lysis, extraction and DNA resuspension and also by the amount of template DNA used in the PCR. It is self-evident that the way of preservation plays an important role as well as the removal of inhibitory substances during DNA extraction, which depends mainly on the employed extraction method. In the case of dermal scrapings on filter paper in the size of about 0.5 cm in diameter all three methods are efficient (phenol-chloroform, guanidine and chelex).

4.9.2. Recommended approach for PCR diagnosis in the future:

The amplification of the ITS-1 region should be the method of first choice. No other system has been found in the literature so far which is simultaneously as sensitive and species specific for the whole spectrum of Leishmania species. It is of great benefit especially for a country like Israel, where 3 species are endemic and any species may be imported. Therefore any Leishmania species can potentially be encountered in Israel. In countries where only one species is known to be endemic it would be sufficient to amplify the ITS-1 region for a genus specific result without consecutive RFLP-analysis. RFLP-analysis could follow if the disease was imported from another country. If a laboratory prefers to concentrate only on genus specific diagnosis, kinetoplast primers 13A/13B should be favoured due to the higher sensitivity. These primers may be employed for the same reason in cases of negativity with the ITS-1 primers. With the advent of the ITS-1 primers the L.braziliensis specific primers MPL1/MP3H can be replaced as well. If the ITS-1 amplification is negative, they may be used in a second attempt (higher sensitivity). The ITS-1 primers are superior in New World leishmaniasis because they can identify also L.mexicana infections, whereas the L.braziliensi s primers can positively identify only species of the L.braziliensis complex.

Kinetoplast primers Uni21/Lmj4 can be replaced by ITS-1 primers due to the higher sensitivity of the latter. There may be still applications where primers Uni21/Lmj4 are benefitial: Whenever a distinction between L.major and L.tropica is required and enzyme digestion (with the ITS-1 method) is not wanted or not possible the kinetoplast primers can be employed. It can be assumed that primers Uni21/Lmj4 could have a high success rate if the best conditions (discussed earlier) are chosen on a regular basis.

4.10. Leishmaniasis in non-endemic areas with particular regard to Germany:

In Germany leishmaniasis is occasionally seen in travellers and immigrants. The incidence of leishmaniasis is not known since the disease is not reportable. Due to tourism the number of [page 105↓] cases is increasing. According to the "Deutsches Ärzteblatt" (Harms-Zwingenberger and Bienzle, 2000) patients are either treated by local physicians or in hospitals. Diagnosis relies on microscopy, culturing, serology and PCR. Rarely cases of infantile kala azar have been diagnosed in the past. One interesting case was reported recently: an infant of 15 months was diagnosed with VL, who had never left Germany except for a short visit to Holland (Bogdan et al , 2001). After 6 months of fever and hepatosplenomegaly kala azar was finally diagnosed. The parents reported the family had spent holidays at a campsite in southern Germany (Füssen) in the previous summer, where many returners from the Mediterranean use to camp on their way back to Germany. One hypothesis suggested that an infected sandfly might have been imported in a car. It was also suggested that autochthonous foci could possibly emerge in central Europe. This is indeed worth-while to consider as a realistic possibility since sandflies have been identified recently in south-west Germany along the upper Rhine valley. The identified sandfly species, Phlebotomus mascittii Grassi 1908, belongs to a subgenus responsible for transmission of VL in the Mediterranean (Naucke and Pesson, 2000). The prospect of a warmer climate in Europe may create conditions for endemic leishmaniasis in central Europe (Kuhn, 1999).

Dogs are much more frequently diagnosed with CVL in Germany than humans with VL. According to Gothe et al ., (1997) 236 dogs were diagnosed with leishmaniasis between 1993 and 1995. The dogs were either imported from Mediterranean countries or had been travelling with their owners. It can not be excluded that infected dogs could potentially introduce L.d.infantum and function also as a reservoir in an emerging focus.

Individual cases have shown, that it is worth-while to search for leishmaniasis in patients with fever of unknown origin (FUO) and splenomegaly, even if the patient history does not suggest kala azar. Unusual ways of transmission have been reported as for example congenital transmission (Meinecke et al ., 1999), blood transfusion (Le Fichoux et al ., 1999) and possibly also liver transplantation (Grimm et al ., 1996). Furthermore, one has to be aware that immunosuppressed persons (eg. HIV-infected persons, transplant recipients, patients with chronic diseases) are at risk to contract VL during visits to Mediteranean countries. It should be also seriously considered to screen transplant organs routinely by PCR if they come from endemic countries in Europe. As transplantation is increasingly organized centrally in Europe this measure could help to prevent cases of “autochthonous VL” in non-endemic European countries. Another interesting fact should be notified: it has been shown by Le Fichoux et al ., (1999) and it has been dicussed also by Kubar et al ., (1997) that the rate of subclinical infections, even with parasitemia, can be considerably high in endemic areas of kala azar (eg. southern France, Nice area). This is a relevant issue with regard to blood transfusion in these areas. Out of 565 blood donors who were screened for the presence of antibodies (by western blot) 76 were [page 106↓] found to be seropositive. The buffy coats of the positive samples were cultured and promastigotes were detected in 9 cultures. Grogl et al , (1993) have studied the survivability of Leishmania parasites in human blood products under blood bank conditions. It was found that the parasites survive the conditions up to at least 25 days and also retain their infectivity. Also tourists to the Mediterranean could contract subclinical infections, and could become symptomatic after years, eg. in a state of immune suppression. The same applies for residents who come from Mediterranean or Middle Eastern countries (eg. Turks, Iranians, Arabs) In non-endemic countries physicians are often not aware of leishmaniases and diagnosis is usually delayed, especially in the serious cases of kala azar with unusual histories. Since infections can theoretically be caused by all Leishmania species the identification at the species level is mandatory. Sensitive direct PCR diagnosis has the same relevancy as in endemic regions and should be favoured as the first diagnostic measure in suspected cases of VL and should be routinely performed in cases of CL in addition to microscopy. Culture based methods and serology can be performed as well but should be second line diagnostic measures. Invasive diagnostic measures, such as skin biopsies (CL) or bone marrow aspirations (VL), can be preserved for the case that the primary PCR on dermal scrapings (CL) or peripheral blood/ buffy coat (VL) is negative and symptoms still support the diagnosis.

4.11. Concluding remarks:

As leishmaniasis is not defeated but is rather a health threat more than ever before, improved diagnosis is of crucial importance for any attempt to control the disease. Diagnostic methods have to be simple and cheap enough to be used also in developing countries, and they have to be suitable for field work too. Endemic areas are not confined to the developing world, as most of the classical tropical diseases are. Also the developed world has to be prepared to see more patients with leishmaniasis, mostly due to endemicity in many countries with a warmer climate, travelling, but also as an opportunistic disease in HIV-infected patients. PCR is a very powerful tool for sensitive and specific diagnosis. The benefits of the PCR-methodology are undoubtedly enormous. This work has focused on PCR-diagnosis of leishmaniasis in Israel and the West Bank, where the methods have been established and are ready now for routine use. This study has been also exemplary for other endemic or non-endemic countries, where direct PCR diagnosis of leishmaniasis is planned to be introduced. The methods can be adopted, benefits of different extraction- and PCR-methods can be estimated with the help of this research and the best combination of methods can be selected for different settings in other countries.


[page 107↓]


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 3.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML generated:
08.12.2003