In this study, 90 samples from 15 microscopically positive P. falciparum patients from southwest Ethiopia were tested for falciparum-specific RNA by QT-NASBA. Microscopy showed mild to moderate parasitaemia in all, and low gametocytaemia in 80% of patients. Median gametocyte density was 7 per 10 μL (range 3-29) corresponding to a density of 0.3-2.9 gametocytes/μL which is well below the usual threshold for microscopy of at least 16 gametocytes per μl . Gametocytaemia was only detected by reading the complete thick smear of 10 μL blood, no gametocytes were detected by routinely reading 200 fields.
The limit of detection for the gametocyte-specific target Pfs16 was 143 copy numbers corresponding to very low submicroscopic densities of gametocytes.
All blood samples were positive for Pfs16-mRNA that is expressed in all gametocyte stages. Only stage I, microscopically indistinguishable from asexual stages, and stage V, the mature gametocytes, are found in peripheral blood in humans. Stages II to IV are sequestered away from circulation . Pfs25-mRNA from mature gametocytes was detectable in 11 patients. The most probable explanation for this is that the other four patients had no mature gametocytes in their peripheral blood yet. Diagnosis of the disease is most often shortly after onset of symptoms, mature stages peak at day 7 after onset of symptoms [25–27], although this is dependent on the treatment. Secondly, the theoretical limit of detection for Pfs25 was 1,710 RNA copy numbers, about ten times less sensitive than Pfs16. However, stage-specific expression levels per gametocyte are not published for Pfs25 or Pfs16, average concentration levels of the specific targets are unknown.
As expected, quantification of copy numbers showed that the density of RNA copy numbers in capillary blood (finger prick) was higher than in blood from the cubital vein . The difference was more prominent for mature gametocytes than for stage I and V together. The chance to be taken up by mosquitoes is higher in small blood vessels and more important for stage V.
Comparison of 25 μL finger prick and 100 μL venous blood showed a higher sensitivity for venous blood but the difference for Pfs25 was negligible. However, Pfs25 was not detectable in a finger-prick sample from one patient where the venous blood sample was positive. Although the density in capillary blood seems to be higher, the amount of blood extracted was only 25 μL. For venous blood, 100 μL blood was extracted. Naturally, the probability to detect any gametocyte increases with the amount of blood extracted if density is very low. As finger-prick blood is more convenient and simpler to get in rural areas with its associated health centres, 25 μL finger-prick blood can be recommended for mature gametocyte screening in epidemiological settings.
Probably all or most of the patients with asexual parasitaemia carry gametocytes but sensitive rapid diagnostic and subsequent elimination of all parasites is essential in symptomatic malaria. However, as transmission reduction is of key importance in eradication efforts, gametocyte-specific diagnostic is needed to control the success of interventions. Gametocyte carriage is age-dependent and decreases with age [25, 29]. Density of gametocytes influences the infection of mosquitoes, very low densities are sufficient but the relation seems non-linear . Investigation of the infectious reservoir includes testing for asymptomatic carriers. This would have been the next step of this study if non-invasive methods had proved more sensitive. Symptomatic malaria patients were chosen for the pilot project as higher gametocyte densities were assumed. To investigate further and reduce the infectious reservoir, finger prick blood spotted on filter paper seems to be a reliable field-proven method but should be validated for asymptomatic carriers.
Gametocyte-specific mRNA detection in saliva, urine and mucosa was low for Pfs16 and negative for Pfs25. Either mature gametocyte mRNA is not secreted at all or the sensitivity of the method is too low. Pfs16 could be found in 20% of urine and 13% of saliva and mucosa samples. Therefore, the detection rate for non-invasive methods is unfortunately insufficient. With the methods used, the fragments amplified were 111 bp for Pfs16-mRNA and 156 bp for Pfs25-mRNA, respectively. It is known that mRNA is fragile and decomposes fast outside of cells. The milieu found during urine filtration and within urine, as well as on mucosal surfaces is especially delicate for extracellular mRNA stability. Therefore, part of the poor detection sensitivity may be due to the decay of the mRNA into smaller pieces in the respective compartment. Under the given circumstances, finger-prick blood collection proved to be the most effective method. Filter paper could be stored at room temperature if it is processed within 28 days. RNA loss would be on average about 20%. If storage duration is longer than one month, RNA loss at room temperature is unacceptably high and the filter paper should be frozen .
Interestingly, mucosa on filter paper performed two to four times better than fresh frozen mucosal swab. The swab seems to conserve the RNA less inefficiently than instant drying and freezing of mucosal swab material on dry filter paper.
A study from The Gambia on DNA of asexual parasites showed a good correlation of saliva parasite density as estimated by quantitative real-time PCR (qPCR) with parasite counts established by microscopy (p = 0.58; P <0.001). Correlation of qPCR results for urine with microscopy counts was poor (p = 0.20; P =0.117). Mean parasite density in blood was about 600-fold and 2,500-fold greater than mean density in saliva and urine samples, respectively. Compared with microscopy results, nested PCR of saliva had a sensitivity of 73% for all samples and a sensitivity of 82% in samples with a parasite density of > or = 1,000 parasites/μL; sensitivity in urine samples was only 32% . PCR-amplicons had a bigger size than amplicons of the NASBA, sensitivity was nevertheless similar. The advantage of RNA is that its detection indicates that there are still living parasites circulating in the blood whereas DNA can be positive several days after acute infection.
The study from Mharakurwa et al. compared blood, saliva and urine samples from individuals with mostly aymptomatic parasitaemia. Samples were spotted on filter paper before extraction. Saliva samples showed a 1.6-fold higher amplification score than corresponding urine samples (p = 0.029). DNA amplicons ranged between 229 bp and 643 bp .
In this study, 18S-rRNA in saliva was detected in 80% of the patients. 67% of urine samples yielded RNA. However, geometric mean parasitaemia (GMP) by microscopy was higher than in the other studies. GMP in The Gambia was 1,785 parasites/μL (CI 95% 695–4,588 parasites/μL) . Furthermore, samples in the Gambian study were processed directly within two hours and not spotted on filter paper before. A recent study about filter paper showed better results for mRNA detection when the blood was spotted on filter paper and dried 24 hours before extraction than direct extraction of fresh blood .
Urine and mucosa tissue proved inefficient for the detection of malaria RNA but saliva analysis as non-invasive method could be conducted in cross-sectional and other studies. Primers amplifying smaller fragments of the target RNA may increase the sensitivity of the test protocol. Saliva could be tested with various diseases including malaria. Again, a rapid test would simplify efforts and allow immediate treatment.