*3.4. Giardia Infection Diagnosis and Treatment*

The first *Giardia* infection was diagnosed by the *Giardia*/*Cryptosporidium* antigen test in Cases 1 and 2. Case 3 was diagnosed by an antigen test and multiplex PCR and Case 4 by cysts stool detection. The detection of reinfection, treatment failure and eradication of *Giardia* were performed by (a) stool cultures and/or the antigen test repetition in Case 2, (b) an antigen test and a multiplex PCR with symptom remission in Case 3, and (c) the remission of symptoms and negative antigen test in Case 4. In Case 1, it was not possible to assure *Giardia* eradication after the first infection because microbiological tests were not performed and the diagnosis was based on the remission of symptoms. In the reinfection, a stool examination was performed and *Giardia* cysts were not detected, confirming its eradication. None of the patients currently have evidence of *Giardia* infection, with antigen test negative.


**Table 1.** Epidemiological, clinical and immunological characteristics of patients with *Giardia* infection and CVID.

NILH: nodular intestinal lymphoid hyperplasia;

 HBV: hepatitis B virus; Ig:

immunoglobulin;

 IVIG: intravenous

immunoglobulin.

First-line *Giardia* infection treatment was metronidazole, which successfully cured infection in Cases 1, 2 and 4. Quinacrine plus paromomycin was the successful treatment used for refractory giardiasis in Case 3 (Table 2).

**Table 2.** *Giardia* infection treatments in each patient.


mg: milligrams; g: grams; h: hours.

## *3.5. Database Review Results*

We found 17 published cases of giardiasis in CVID. The median age was 39 years old (15–62), eleven were male (64.7%) and six were female (35.3%). Six were refractory giardiasis (35.3%). The main clinical manifestations were diarrhea (14/17; 82.3%), weight loss and abdominal symptoms (9/17; 52.9%), and three patients had splenomegaly (3/17; 17.6%). All CVID patients showed undetectable or decreased IgA levels. All summarized data are collected in Table 3.


**Table 3.** Cases of giardiasis in CVID

documented

 in the literature.

#### **4. Discussion**

The incidence of symptomatic giardiasis in our CVID cohort was 12.9%. The main clinical manifestations of *Giardia* infection were diarrhea, weight loss and abdominal pain, both in our patients and in review patients. It is important to make a correct differential diagnosis with other entities such as celiac disease and inflammatory bowel disease (IBD), which was performed in our patients, because gastrointestinal symptoms are very common in CVID, especially transient or persistent diarrhea [35]. Gastrointestinal *Giardia* symptoms in Cases 2 and 4 produced protein loss and malabsorption, which made it difficult to maintain their IgG levels in the normal range despite IVIG treatment. Immunoglobulin replacement therapy is the basic treatment of these patients, which improve their symptoms and their quality of life, but also prophylactic and therapeutic antibiotics for their recurrent infections [8]. Antibiotics stimulate intestinal dysbiosis, producing digestive alterations and favoring chronic or refractory infections by gastrointestinal pathogens such as *Giardia*. The gut microbiome has been shown to play a key role in determining susceptibility or resistance to *Giardia* colonization by modulating immune responses, while the parasite itself can influence the immune response to the host [36]. Recent research has recognized giardiasis as an important risk factor for developing long-term postinfectious syndromes, such as IBD, chronic fatigue syndrome, and extraintestinal consequences such as arthritis or allergies, even months or years after parasite clearance [16,37], which could be one of the reasons for the Crohn's disease-*like* symptomatology in Case 4.

It should be noted that Cases 3 and 4 had nodular intestinal lymphoid hyperplasia (NILH), which is also found in CVID patients evaluated in the literature review (6/17; 35.3%). NILH is a rare condition that can occur as a form of compensation for the functionally inadequate intestinal lymphoid tissue found in patients with immunodeficiencies. However, NILH has also been associated with *Giardia lamblia* infection, and in some cases, its eradication has resulted in NILH reduction [23], or with an overregulation of the response mechanisms of the lymphoid tissue associated with the digestive tract [29]. In Cases 3 and 4, neither possibility can be ruled out, but it would be interesting to monitor its evolution because NILH could represent an intermediate stage and a risk factor for the development of lymphoma [38].

IgA, Th17 and CD4+ T cells are key in the immune response against *Giardia* [39]. Undetectable or very low levels of IgA are a characteristic in our four patients and in all available review patients, an immunoglobulin that is not replaced with IVIG treatment and that is key in the defense of the intestinal mucosa. Various studies have associated its deficiency with a greater capacity for binding and proliferation of foreign pathogens such as *Giardia* to the intestinal epithelium [12]. T CD4+ lymphopenia found in Case 3 could be another immunological factor resulting in refractory giardiasis. The remarkable reduction of B cells in Case 1 led us to suspect Bruton's disease but it was ruled out by the NGS study. A remarkable NK cell lymphopenia was detected in Cases 1 and 4. These cells are important in the antiviral and antitumor defense. However, in CVID patients, NK cells have been associated with high frequencies of severe bacterial infections and non-infectious complications, suggesting that NK cells also have a role in controlling bacterial infections [40].

The EURO-Class classification was a multicenter European effort that allowed defining different subgroups of patients with CVID based on their B cell phenotype and some clinical features [6]. In our study, all cases had a reduction in switched memory B cells. The increase in CD21*low* cells is the strongest marker associated with splenomegaly. Case 3 had splenomegaly, but CD21*low* B cells were not altered, and she had also a remarkable increase of transitional B cells, without lymphadenopathy. In one literature case with refractory giardiasis, lymphadenopathy and splenomegaly were detected, but both regressed after *Giardia* eradication [26].

Different genetic variants have been involved in the pathogenesis of CVID, but most patients do not have a specific genetic defect background [41]. The NGS analysis detected possible CVID-associated variants in two of our patients. In Case 2, a heterozygous variant in *TCF3* gene (p.Lys519Glu) was detected. This gene is located on chromosome 19p13.3 and codes for transcription factors that promote the expression of genes involved in lymphopoiesis, differentiation and maturation of B and T cells. This variant has not been previously described but its presence in the control population and in the asymptomatic carriers of families with severe agammaglobulinemia caused by biallelic variants, as well as the contradictory results of in silico predictors, point to a benign variant [42,43]. In Case 4, a heterozygous variant in *PTEN* gene (p.Val365Ile) was detected. This gene is located on chromosome 10q23.31 and codes for a phosphatase that acts as a tumor suppressor gene. This enzyme also participates in the PI3K/Akt signaling pathway and, thus, in the differentiation and homeostasis of T and B cells. It has previously been considered a variant of uncertain significance (VUS) according to human databases in patients with Cowden syndrome type 1, a condition within the group of syndromes related to the development of hamartomas (PHTS) following an autosomal dominant pattern of inheritance [44]. These patients had heterogeneous phenotypes, some of whom had PHTS with a CVID phenotype, exhibiting a decrease in switched memory B cells and a reduction in functional and mature NK cells, immunological alterations that are present in our Case 4 patient. Likewise, lossof-function mutations in *PTEN* have also been described as being associated with activated phosphoinositide-3-kinase delta syndrome-*like* (APDS-*like*) because it acts as an antagonist in the PI3K-delta signaling pathway [45,46]. Cosegregation and functional studies will be necessary to confirm the pathogenicity of this variant in the context of CVID, which will be a future objective of our group.

The HLA class I and II genes code for cell surface molecules specialized in antigen presentation to T cells and play a key role in the immune response. Previous studies have reported a positive association between certain HLA alleles and different diseases, including susceptibility to CVID [47]. It has been described that the extended haplotype HLA-A1, -B8, -DR3 is more frequent in these patients [48] and that homozygosity in the HLA class II region, specially HLA-DQ, is associated with greater susceptibility to CVID [49]. Curiously, Case 3 patient has homozygosity in HLA class II, our giardiasis refractory case. This homozygosity could result in a lower repertoire of HLA class II molecules on the cell surface, resulting in a lower capacity to respond to foreign or pathogenic antigens and contributed to CVID susceptibility to specific environmental conditions. Moreover, Case 4 has the HLA-DRB1\*13:01 allele, which has been previously associated with susceptibility to *Giardia* infection [50]. On the other hand, Case 1 patient has the C\*06:02 allele, which is one of the most strongly HLA alleles associated with psoriasis susceptibility [51]. None of the four patients presented the ancestral haplotype 8.1 (HLA-A1, -B8, -DR3), although Case 1 had the ancestral haplotype 44.3 (A\*29:02; B\*44:03; C\*16:01; DRB1\*07:01; DQB1\*02:02). It would be advisable to extend the study to a bigger CVID cohort to have more precise data on this evidence.

First-line giardiasis treatment is based on the use of nitroimidazoles [11]. In our cohort, 75% (3/4) responded to metronidazole, evidenced by negative microbiological tests and remission of symptoms. This is similar to the case reports included in the literature review, although reinfections occurred in all of our cases. However, one of the highlights of this work is the first refractory giardiasis case diagnosis in our hospital in a CVID patient (Case 4) and, to our knowledge, the first time that the combination of quinacrine plus paromomycin has been described and has been effective. We consider the treatment successful because stool examination has remained negative to date. So far, only six refractory giardiasis cases in CVID patients have been published, and different curative treatments have been applied [20,22,26,30,31,33]. Currently, there is no standard therapy for refractory giardiasis, and multiple drugs have been tried [15]. Tinidazole, a derivative of metronidazole, has been shown to be more effective than metronidazole [11,39] but was not effective in Case 3. The combination of metronidazole and albendazole, the latter a benzimidazole, is more effective in treating refractory disease than albendazole alone [15], but again no response was obtained as a third-line treatment. On the other hand, the efficacy and safety of the use of quinacrine in refractory giardiasis have been demonstrated in various studies, mainly in monotherapy or combined with metronidazole [22,52,53]. Quinacrine, also known as mepacrine, was the first antimalarial drug used to treat giardiasis, but its use was limited by detecting possible adverse effects of psychosis. Despite everything, it is usually well-tolerated and has a clinical efficacy of >90% [54]. In the case of paromomycin, its activity is variable against *Giardia* (55–90%) and is rarely used. It can be employed in cases of resistance or pregnancy because it has a low absorption spectrum and does not have systemic effects [14]. It is usually well-tolerated, although being an aminoglycoside, it can cause nephrotoxicity and ototoxicity [15,54]. The quinacrine and paromomycin combination therapy should be used in a larger patient cohort to confirm its efficacy and safety, and more clinical trials are needed to establish the optimal therapy for patients with refractory giardiasis.

*Giardia* drug resistance is currently undetectable in most microbiology laboratories. This is due, on the one hand, because the parasite culture is difficult and takes a long time, and success rates are relatively low [15] and, on the other hand, because resistance depends more on epigenetic factors and post-translational modifications than on genetic variants in the parasite genome. For these reasons, it is not yet possible to develop a routine microbiology technique to detect *Giardia* susceptibility [14]. Moreover, since there are few cases of refractory giardiasis, multicenter studies are required to define the best therapeutic alternatives. At the diagnostic level, it would be interesting to develop a strategy that combines the culture of the parasite with sequencing and comparative proteomics [55], which may allow moving from empirical treatment to targeted treatment in the future.

The early detection and treatment of infections in patients with CVID are crucial to ensure a good quality of life. Moreover, the immunological and genetic characterization of these patients is necessary to improve their clinical follow-up and to better understand the pathogenesis of CVID. One limitation of this work was that the cohort of CVID patients who suffered from symptomatic *Giardia* infection was small. Still, to our knowledge, we present a probably pathogenic *PTEN* variant associated with the CVID phenotype, and, for the first time, a case of refractory giardiasis in CVID that was successfully treated with quinacrine plus paromomycin combination. These data could contribute to the decision-making and therapeutic management of future patients with CVID and giardiasis, especially in refractory giardiasis cases.

**Supplementary Materials:** The following supporting information can be downloaded at https://www. mdpi.com/article/10.3390/jcm11237007/s1, Supplementary Material S1: Genes associated with primary immunodeficiency analyzed by Next-Generation sequencing; Supplementary Tables S1 and S2: Human Leukocyte Antigen (HLA) genomic typing of CVID patients with giardiasis; Supplementary Material S2: Genetic results and sanger sequencing methodology.

**Author Contributions:** Conceptualization, M.Á.L.-N. and C.H.-T.; formal analysis, I.D.-A., J.M.G.- A., P.J. and M.Á.L.-N.; funding acquisition, M.Á.L.-N.; investigation, I.D.-A., J.F.G.-B. and A.E.-O.; methodology, I.D.-A. and M.Á.L.-N.; supervision, M.Á.L.-N., C.H.-T. and P.A.; visualization, I.D.-A., J.F.G.-B., A.E.-O. and J.M.G.-A.; writing—original draft preparation, I.D.-A., J.F.G.-B.,A.E.-O. and P.A.; writing—review and editing, I.D.-A., J.F.G.-B., A.E.-O., M.Á.L.-N. and C.H.-T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was partially financed by Palex Medical S.A.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki, reviewed and approved by the Portal de Ética de la Investigación Biomédica de Andalucía, PEIBA (Cod. 1206-N-22).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** This study is part of the doctoral thesis of Irene Díaz-Alberola, within the program of Biomedicine, conducted at the University of Granada, Spain.

**Conflicts of Interest:** J.M.G.-A. is an employee of Health in Code S. L. All authors declare no conflict of interest.
