Clinical characterization of acute and convalescent illness of confirmed chikungunya cases from Chiapas, S. Mexico: A cross sectional study

Background The emerging chikungunya virus (CHIKV), is an arbovirus causing intense outbreaks in North America. The situation in Mexico is alarming, and CHIKV threatens to spread further throughout North America. Clinical and biological features of CHIKF outbreaks in Mexico have not been well described; thus, we conducted a cross sectional study of a CHIKV outbreak in Chiapas, Southern Mexico to further characterize these features. Methodology/Principal findings We collected blood samples from patients suspected of having chikungunya fever (CHIKF) who presented to Clinical Hospital ISSSTE Dr. Roberto Nettel in Tapachula, Chiapas, Mexico. In addition to the clinical examination, real-time polymerase chain reaction (PCR) standardized for the Asian Chikungunya lineage and/or enzyme-linked immunosorbent assay for immunoglobulin M (IgM) were used to confirm CHIKV diagnosis. Of a total of 850 patients who presented with probably CHIKV at Hospital “Dr. Roberto Nettel”, 112 probable CHIKF cases were enrolled in this study from November 2014- June 2015, of which 95 patients (84.8%) were CHIKV positive and 17 were negative (15.2%). Of these 95 CHIKV positive patients, 62 were positive by real-time reverse transcriptase PCR (+qRT-PCR); and 33 were seropositive to +IgM with a negative qRT-PCR. The most frequent symptoms reported were fever (100%), headache (82.3%), polyarthralgia (72.1%), and exanthem (82.3%). Biological abnormalities observed during CHIKV infection were lymphopenia (41.1%), leukopenia (51.6%), elevated transaminases (30.5%-46.3%) and high LDH (46.3%) and CRP (60.0%). Conclusion Clinical and biological data obtained from this study is providing more useful information for benchmarking purposes with outbreaks from different parts of the world and would be helpful for better patient care and treatment.

Introduction evolved from the Eastern/Central/Southern African (ECSA) genotype, was recently obtained using a mouse infection model [23].
The clinical presentation good outcomes based on genotype/lineage have been poorly discussed. This study is an attempt to document and describe the clinical and biological features of CHIKV outbreak caused by the Caribbean linage in Mexico. In addition, we compared clinical descriptions of outbreaks caused by the ECSA-IOL and Asian genotype worldwide with the results obtained in our study. We conducted a cross-sectional observational study in a secondary healthcare hospital in Tapachula, Chiapas, Mexico, at which probable CHIKV cases were recruited and then diagnosed by molecular and serological testing. Additionally, clinical symptomatology was recorded by each enrolled patient and then blood cell count and biochemical analysis were performed to obtained relevant biological data related to CHIKV infection.

Ethical considerations
The study was approved by the Bioethical Committee of the National Institute of Public Health (Project #1312 and approval #1738). Patients were enrolled in the study during a CHIKV outbreak, and blood samples were collected as part of the routine laboratory analysis of patients admitted to Hospital Dr. Roberto Nettel (S1 File). Therefore, only an oral consent to participate in the study was obtained from each patient. All pertinent information about the study (purpose, procedures, risks, benefits, alternatives to participation, etc.) was explained to potential participants in the presence of an independent witness, and every patient who agreed to participate was recorded in an intake form by the hospital's Epidemiology Department [26].

Study population and recruitment
The Hospital "Dr. Roberto Nettel" attends only State workers and their immediate relatives who represent 4.5% of the total population of Soconusco Region [27]. Patients who were admitted in this hospital were included in the study if their clinical symptoms met the following inclusion criteria for suspicion of CHIKF: acute onset of fever >38.5˚C accompanied by severe arthralgia not explained by other medical conditions in patients residing in the epidemic area [28]. Exclusion criteria were if patients disagree to participate in the study or were DENV-positive by RT-PCR or ELISA. polyarthralgia, arthritis, headache, myalgia, edema, exanthema, adenopathies, retroorbital pain, digestive alterations (vomiting, diarrhea, and nausea), etcetera. To avoid any confusion between polyarthralgia and arthritis, we differentiated them being the first a transient, intermittent, or persistent pain in multiple joints while the second is the inflammation of one or more joints. Additional epidemiological data such as patient age, gender and travel history is also included. After the clinical examination, the physician filled the format by patient and then collected three blood samples from each patient: one tube with EDTA for a complete blood cell count and one tube with citrate and one dry-tube for the remaining laboratory tests. Dry-tubes were centrifuged at 3,000 rpm, and serum was aliquoted and stored at −80˚C for testing.

Diagnostic qRT-PCR and serology
RNA was extracted from 140μL of serum from each sample using the QIAmp 1 Viral RNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. For molecular diagnostics, one-step qRT-PCR was performed according to the Center of Disease Control (CDC) protocol using primers and probe designed for Caribbean lineage (Asian genotype) [29]: 1) we used the primers CHIK 856 5'-ACCATCGGTGTTCCATCTAAAG-3' and CHIK 962c 5'-GCCTGGGCTCATCGTTATT-3' and the TaqMan Probe CHIK 908 5'-FAM-ACAGTGGTTTCGTGTGAGGGCTAC-NFQ-3´; 2) the qRT-PCR was prepared with the QIA-GEN QuantiTect Probe RT-PCR kit 1 (Qiagen, Venlo, Netherlands) in accordance with the manufacturer's protocol, adding 10μL of RNA sample per reaction in duplicate; and 3) the samples were run in the 7500 Fast Real Time PCR System (Applied Biosystems, Foster City, CA). Thermal cycling conditions were: one cycle of 50˚C for 30 minutes and 95˚C for 15 minutes followed by 45 cycles of 95˚C for 15 seconds and 60˚C for 60 seconds. The cycle threshold (CT) value was 38: results were considered positive if less than 38 cycles were needed in both wells; negative if greater than 38 cycles were needed in both wells, and equivocal if less than 38 cycles were needed in one of the two wells. Serum samples were screened for anti-CHIKV IgM antibodies by enzyme-linked immunosorbent assay (ELISA) as previously described [30,31] using the CHIKjj Detect MAC-ELISA kit (InBios, Inc., Seattle, WA). All samples were tested in duplicate and any inconclusive samples were retested.
For dengue diagnosis, the samples were separately sent to the State Laboratory of Public Health of Chiapas. Serum was tested with Panbio 1 Dengue Early ELISA and Panbio 1 Dengue IgM Capture ELISA (Alere, Waltham, MA) for detection of NS1 antigen and IgM antibody following the manufacturer's protocol.

Results
Ministry of Health reported 2,500 probable chikungunya cases distributed in Soconusco Region from November 2014 to June 2015, of which 850 were admitted by Hospital "Dr. Roberto Nettel". Of these patients, only 112 agreed to enrollment in this study (Fig 1). Complete clinical, epidemiological, and biochemical data were obtained from each participant; and none was excluded because of missing data. Ninety-five patients had confirmed CHIKF with 62 being +qRT-PCR and 33 +IgM/-qRT-PCR, while the remaining 17 patients were CHIKVnegative (Fig 1). Only one patient was qRT-PCR and IgM positive and was categorized as a viremic case in subsequent data analysis. None of the confirmed CHIKV cases was DENVpositive by NS1 or IgM ELISA. Confirmed CHIKV cases from this study were distributed across nine municipalities, with the greatest number of cases in Tapachula (39 cases, 41.0%), Tuxtla Chico (17 cases, 17.9%), Huehuetan (17 cases, 17.9%), and Cacahoatan (11 cases, 11.6%). Data are shown only for viremic and post-viremic cases because the numbers of negative cases were too few (17 cases) for meaningful analysis.
Demographic data for the 95 patients with confirmed CHIKV are shown in Table 1. The mean age of patients was similar at 37.2 and 39.2 years for viremic and post-viremic patients, respectively. The age groups most affected by CHIKV infection were adults 31-40 and 41-50 years, together accounting for almost half of the CHIKV cases in both phases of disease; however, minors (<20 years) accounted for 12.6% of confirmed cases in both stages. Furthermore, sex distribution was unequal showing a male/female rate of 0.4. Approximately 70% of patients in both phases of disease were female (Table 1).
An extensive clinical description of the patients during the viremic and post-viremic phases is described in Table 2. The timing of clinical data and blood collection was an important issue considered in the study. The median numbers of days between symptom onset and consultation (SOC) were 3.0 and 4.0 for viremic and post-viremic patients, respectively. Despite this short time difference in SOC between both phases, we observed some interesting differences in clinical and laboratory features between the disease phases. The most common symptoms recorded during the viremic stage were fever (100%), polyarthralgia (100.0%), headache (82.3%), exanthem (82.3%), chills (82.3%), and nausea (71.0%). Other symptoms often observed in the viremic patients included pruritus (66.1%), arthritis (58.1%), abdominal pain (46.8%), edema (43.5%) and diarrhea (33.9%). On the other hand, the most frequent symptoms described in post-viremic patients were polyarthalgia (100%), headache (87.9%), exanthema (81.8%), chills (75.8%), and nausea (48.5%) and other symptoms developed, such as arthritis (78.8%) and adenopathy (60.6%). The only symptoms that showed substantial differences between phases were arthritis, adenopathy, and nausea ( Table 2).
Regarding laboratory studies, serum samples were tested for different features: blood cell count, carbohydrates, lipids, electrolytes, inflammatory marker, and hepatic and renal functions. These tests were performed to find any alteration caused by CHIKV infection in the viremic and post-viremic patients, and the results are shown in Figs 2 and 3. Most parameters measured in this study fell within normal ranges in both phases except for hepatic enzymes (AST, ALT, ALP), absolute count of white cells, lymphocytes, LDH and CRP. Of these features, however, notable differences between the two phases were observed in lymphocyte count, LDH and CRP levels. Although differences between the phases were also observed in triglycerides, PT, and PTT, these results were not relevant because most of the values were within normal ranges. Table 3 demonstrates the relevant abnormalities in laboratory values with the respective reference values included. Elevated concentration of hepatic enzymes (AST, ALT, ALP) was seen in approximately 30.6-48.4% and 24.4-42.4% of viremic and postviremic patients, respectively. Low white cell counts occurred in 60-70% of cases in both disease stages, and lymphopenia was significantly more frequent in viremic patients (56.6%) than in post-viremic patients (12.1%). Moreover, LDH levels were significantly more frequent in post-viremic patients (69.7%) than in viremic patients (33.9%). CRP was expectedly elevated in the viremic patients of infection (61.0%) but normal in all post-viremic patients.

Discussion
At the end of 2014, chikungunya virus emerged in Mexico and, since then, have caused outbreaks in the states along the Pacific and Gulf of Mexico coasts. Previous sequencing data Clinical description of CHIKF in Mexico suggests that the CHIKV entered the country through the Mexico-Guatemala border in Chiapas [4]. Precisely in this point of entry, we conducted clinical and laboratory evaluation of 112 patients with symptoms concerning for CHIKF, confirming 95 cases by qRT-PCR and ELISA. Depending on whether the patient was qRT-PCR or IgM positive, the patient was placed into one of two groups, the viremic patients group or the post-viremic patients group. From these patients, we analyzed data from the clinical data and laboratory tests; described basic demographic information, such as age and gender.
The key results were as follows: 1) of 112 probable CHIKF cases, 95 cases (84.8%) were confirmed by qRT-PCR (33) or IgM ELISA (62) being adults and females more affected; 2) fever, polyarthralgia, exanthem, and headache were the most common symptoms observed in patients, however, adenopathies that had not been deeply described in previous studies were also found in our study, 3) in the disease onset, symptomatology between previous IOL outbreaks and Mexican outbreak caused by Asian lineage seems to be similar and 4) biological abnormalities observed during CHIKV infection were lymphopenia, leukopenia, elevated transaminases and high LDH and CRP. In comparing our results to previously report epidemiological data from other outbreaks, six of nine studies found that the mean or median of age of infected patients was between 30-40 years, which was similar to the mean age of 37.9 years in our study. Most infected patients in most other studies were adults between 20-40 years with percentages between 30 to 50%, with lower frequencies in the elderly and children such as our study. However, two outbreaks from La Reunion Island reported that elderly people between 50-60 years were most affected. Lack of herd immunity in the region may explain why CHIKV infections occurred more often in adults than in old and young patients [32].
The male-to-female ratio in confirmed CHIKV cases has varied. While Thiberville et al.   2015) each had low numbers of study participants, suggesting that their results are not representative of the general population. The sample obtained for this study is also non-representative and male-to-female ratio of total of probable cases (850 cases) was 0.5 while the ratio of enrolled participants was 0.4 (112 cases). However,  it is important to consider that CHIKV transmission might have occurred more in home-environments putting housewives at major risk of infection, such as was described in previous reports of dengue seroprevalence in Mexico [33]. The most common symptoms reported by all previous studies including ours were fever (80-100%), polyarthralgia (80-100%), headache (50-80%), rash (30-70%), and digestive alterations (20-80%). Adenopathy, arthritis, and edema were rarely described in these studies. It is important to point out that exanthem was very high (82.3%) in our study compared with the rest of studies that is seen up to 50% of patients [34].
Cervical lymphadenopathies were observed in Indian CHIKV outbreaks and imported Spanish cases as well as in patients affected by the O'nyong-nyong virus, a related alphavirus [35][36][37]. Interestingly, Staikowsky et al. (2009) observed an increase of adenopathies from 8.9% in the viremic patients to 29.0% in the post-viremic patients of CHIKV, similar to our study from 33.3% in the viremic phase to 60.6% in the post-viremic phase. The occurrence of lymphadenopathy possibly may have been underdiagnosed in the rest of studies, given that it may occur toward the end of the febrile period, during recovery from the infection, and appears to be self-limiting [37]. Unfortunately, the specific location of adenopathies has not been not recorded in most of studies and neither have we in this study. Regarding arthritis and edema, the lack of sufficient data makes determining the prevalence of these symptoms difficult.
Data including viral genotype from the La Reunion Island, Singapore, Gabon, Sri Lanka, Colombia, and Trinidad & Tobago outbreaks are included in Table 4. Although the Caribbean Asian lineage of CHIKV has been reported as less virulent than the IOL lineage [23], the Asian lineage outbreak from Chiapas and the rest of the ECSA-diverged IOL outbreaks worldwide appear to behave clinically similarly during the first 3-4 days of infection.
Hepatic function was evaluated in this study by measuring the levels of AST, ALT, and ALP. Studies of the outbreaks in La Reunion Island and Gabon reported elevations of these enzymes in approximately 20-46% of cases with AST levels up to 55 and 53.2 U/L in viremic and post-viremic patients, respectively. Our study observed elevations of these enzymes in approximately 30-50% and 24-43% of viremic and post-viremic cases, respectively. Our results are quite similar to those previously reported in the literature, but this increase may be explained by the use of paracetamol and/or ibuprofen for the treatment of symptoms, producing transient drug toxicity in the liver [11]. Another biological feature that has been associated with CHIKV infection is elevated C-reactive protein. Previous studies from the outbreak in La Reunion Island reported high levels of CRP in approximately 50-70% of CHIKF cases, with serum levels up to 5.6 mg/L, similar to the elevated levels of CRP we found in 61.3% of patients disease. Additionally, another enzyme observed raised was LDH in post-viremic patients. This was previously observed elsewhere [38,39] and explained because this enzyme is released into the extracellular environment during cellular injury associated with inflammation [40].
The blood cell count parameters that were abnormally decreased during CHIKV infection were absolute counts of leukocytes and lymphocytes, which occurred in 54.8% and 56.5% of cases, respectively. In post-viremic cases, leukopenia persisted in 45.5% while lymphopenia decreased to 12.1% in the post-viremic patients. Only lymphopenia was reported in 70-80% of confirmed cases in the outbreaks of La Reunion Island, with counts of 608 cells/mm 3 during the viremic phase. Leukocytosis had been associated with CHIKV infections in the past [41,42]; however, none of the studies cited here found a significant increase in white cells. Moreover, a significant decrease of these cells was observed in Win et al. (2010) and our study.
Limitations of our study must be considered. We obtained a biased sampling, including only patients for one hospital because of the ease of obtaining samples from its clinics, thus, not allowing for a sample that was truly representative of the general population. Another Most of previous outbreaks have also faced similar limitations with sample collection, but they were able to undertake meaningful evaluation of the clinical information such as we were able too.
In conclusion, we describe the epidemiology and clinical features of CHIKV infection in Southern Mexico. Clinical data of different outbreaks worldwide would be helpful for the investigation of potential biomarkers for each of the phases of chikungunya fever and therefore, clinical diagnosis and treatment would be improved in viremic and immediate post-viremic phase.
Supporting information S1 File. Explanation of patient attention during the CHIKV outbreak in the hospital "Dr. Roberto Nettel" of 2014-2015. This appendix clarifies how patients were asked for an oral consent to provide blood samples keeping confidentiality of them as part of routine process of epidemiological surveillance of febrile diseases outbreaks. (PDF)