To our knowledge this is the first report of plasma and intracellular levels (in PMN und PBMC) of AZM in MALT lymphoma patients undergoing continuous oral treatment for 6 months. We found the mean plasma concentrations (58.97 ng/ml, SD ±30.48) observed in our patients to be relatively high and stable over the whole period of application (Figure 1a). Yet at week 24 AZM levels appeared to be slightly decreased. However, this fact should be interpreted with caution as we lost follow-up of 6 patients at week 24 and plasma levels used to be above average in these patients.
A study by Amsden et al. was the only one using 1500 mg AZM as a single shot therapy in healthy volunteers (28). After 7 days they reported mean plasma levels of about 30-40 ng/ml. Other study groups used different dosage regimes: Matzneller at al. applied AZM 500 mg for 3 days and plasma concentrations determined 5 and 10 days after the beginning of treatment were 27.8 and 8.15 ng/ml (29). Olsen et al. used 500 mg once, followed by 4 daily doses of 250 mg AZM. 7 days after the last dose they reported plasma concentrations of 15 ng/ml (30).
In accordance with these previous studies on healthy individuals (28-30), we conclude that a dosage regime of 1500 mg AZM once weekly in patients with MALT-lymphoma provides continuously high blood levels and, as AZM is distributed quickly from plasma to various tissues and leukocytes with considerable accumulation, correspondingly high concentrations in these compartments.
In order optimize future treatment approaches, we were interested in potential influences on azithromycin distribution in blood plasma, such as age, body weight and BMI. Due to its rapid redistribution from the vascular compartment to tissues and leukocytes, the plasma concentrations of AZM after several days are mainly a result of slow efflux from these compartments rather than biliary or renal elimination. Thus, it can be suspected that individual distributional differences of adipose, muscular and lymphatic tissue as well the extent of local inflammation with associated cell accumulation are the main constituents of AZM plasma concentrations and that these parameters are not sufficiently represented by measuring weight and BMI alone.
One of our patients has shown particularly high plasma through levels which were more than three standard deviations above the mean. In fact, this patient had the lowest body weight with 41,0 kg, which might explain the alteration. Still, we were unable to find any significant correlation between body weight and AZM plasma concentrations (Table 1).
Aging is typically accompanied by an increase of fat tissue, a relative reduction in muscle mass, as well as a decrease in glomerular filtration rate (31). In our patients with an age ranging from 47-88 years and a mean age of 66,5 years these influences might be of particular relevance.
However, in respect of clinical implications we found plasma concentrations to be independent of the factors we examined and therefore an adaptation of AZM dosing seems unnecessary. Of course, one has to take into account that one limitation of our study was the rather small study population and therefore there might have been a lack of statistical power to expose clinically relevant differences. We are aware, that the concentrations we measured in plasma and intracellularly show considerable variance. This might be explained by the uncertainty of measurement which would expected to be more prominent when working with very low concentrations. Assessing AZM levels by HPLC methods, however, has proven to be reliable even at very low concentrations (32).
Regarding intracellular levels of AZM, we measured higher drug levels in PMN than in PBMC (39274 vs. 6648 ng/ml) (Figure 3, 4). These results are not in line with those of Amsden et al., who measured higher concentrations in PBMC than PMN (27 vs. 18 mg/l) 10 days after application of 1500 mg AZM (28). Wildfeuer et al., on the other hand, reported 53 mg/l in PMN and 1 mg/l in PBMC 14 days after an administration of 500 mg AZM for 3 consecutive days. However, none of those patients had undergone continuous dosing, and different methods in the handling of the blood samples could also be an explanation. In fact, Wildfeuer et al. described losses of about 25% of AZM from monocytes during one step of washing with PBS. Furthermore, when calculating intracellular concentrations, different study groups used different cell volumes (e.g. Wildfeuer et al. for PMN 450 fl; Amsden et al. for PMN 334 fl) (33). That indicates that the measured concentrations are not directly comparable with those of other study groups.
Furthermore, the reported studies were performed on healthy subjects whereas pathophysiological differences in lymphoma patients might contribute to the pattern of AZM distribution in this group. MALT lymphoma is characterized by chronic inflammation (34). Due to azithromycin being a weak base it accumulates at sites of low pH, such as tissues with ongoing inflammation and intracellularly in lysosomes and phagosomes (35). Our study population involved patients with MALT lymphomas of different origin, stage and disease activity. Taken into account that MALT lymphoma appears as a systemic disease in a multitude of cases, these factors might be, at least partly, an explanation for the heterogeneity of our data (36).
Our expectation was, that as concentrations in PBMC might be similar to those of MALT lymphoma-lymphocytes, their determination could show a link between AZM dosing and treatment outcomes. Specifically, we were interested if failure to treatment might be due to insufficient AZM concentrations in lymphocytes. Our data did not hint at such a connection. However, in vitro studies by Ratzinger et al. show that relevant mTOR inhibitory concentrations require AZM concentrations of 10 mg/l in the culture medium, indicating far higher intracellular levels in lymphocytes under these conditions (21). Achieving such levels in vivo reasonably and safely is improbable. Nevertheless, further studies on AZM effects on MALT lymphomas using higher doses might be promising. It has been shown that even high daily doses (15 mg/kg/day over 24 weeks) of AZM in patients treated for cystic fibrosis were well tolerated (37). Apart from those data, we think that determining AZM concentrations in biopsy material of MALT lesions might be the most direct way to assess the situation on site. One could then evaluate, whether the local and the intracellular concentrations in PBMC and PMN are corresponding.
Interestingly, the AZM levels in PMN were significantly higher in patients without a response to treatment (Figure 4). This result has to be taken with some caution, as we were only able to compare 4 patients responding to therapy with 11 non-responders.
In case of gastric MALT-lymphoma, it is known that the ongoing inflammation attracts and activates neutrophils. Patients who were unresponsive to AZM therapy had either stable or progressing disease, indicating a greater disease activity and level of inflammation in this subgroup. Due to the pharmacological properties of AZM mentioned before, one can suspect an accumulation of AZM in PMN based on the increased disease activity in these patients, which would explain why this patient subgroup suffered a poor outcome.