Recovery after more than one month of dialysis in legionella-associated acute kidney injury: a case evaluated with gallium-67 scintigraphy and renal biopsy

Abstract

Distinguishing tubulointerstitial nephritis (TIN) from acute tubular necrosis (ATN) is clinically important in Legionella-associated acute kidney injury (AKI), because TIN may respond to corticosteroid therapy whereas ATN does not. However, kidney biopsy is often not feasible in critically ill patients, and noninvasive approaches to guide treatment decisions are needed. We report a 73-year-old man with severe Legionella pneumonia complicated by dialysis-dependent AKI persisting for more than one month despite resolution of the infection and discontinuation of all antibiotics by Day 9. Markedly elevated urinary beta2-microglobulin (118,539 μg/L at Day 28) and diffuse intense bilateral renal uptake on gallium-67 scintigraphy (grade 5, clearly exceeding liver and vertebral body uptake) on Day 31 supported ongoing renal inflammation in the setting of prolonged tubular injury. A diagnostic trial of prednisolone (0.5 mg/kg/day) was initiated; from approximately 1 week later, urine output increased, serum creatinine declined, and hemodialysis was successfully discontinued on Day 43. Renal function returned to pre-illness baseline by Day 57 (serum creatinine 0.75 mg/dL). Kidney biopsy performed after clinical stabilization showed acute tubular injury with chronic tubulointerstitial changes and a focal, reactive T-cell infiltrate; prior or partially treated TIN could not be excluded. This case shows that renal recovery remains possible in Legionella-associated AKI even after more than 1 month of dialysis dependence, and suggests that gallium-67 scintigraphy, interpreted alongside serial urinary tubular injury markers, may provide ancillary evidence of ongoing renal inflammation when biopsy is not immediately feasible.

Introduction

Legionella pneumonia can progress rapidly to severe respiratory failure. In this setting, acute kidney injury (AKI) is a frequent complication with important prognostic implications. A recent cohort of mechanically ventilated patients with Legionnaires’ disease reported AKI in 74% of cases, with 44% requiring kidney replacement therapy [1].

Although rhabdomyolysis and acute tubular necrosis (ATN) are well-recognized mechanisms of kidney injury in Legionella infection, tubulointerstitial nephritis (TIN) is another important and potentially treatable cause. Distinguishing TIN from ATN has direct therapeutic implications: TIN may respond to corticosteroid therapy, whereas corticosteroid use in isolated ATN offers no benefit and may increase the risk of infection. In critically ill patients, however, renal biopsy is often not feasible because of hemodynamic instability or other procedural risks. In such cases, noninvasive tools that help characterize ongoing renal injury become particularly important. Gallium-67 scintigraphy reflects active inflammatory infiltration of the kidney and has been reported to favor TIN over ATN when diffuse intense bilateral uptake is present [2, 3], but its role in guiding treatment decisions in Legionella-associated AKI has not been well described.

We report a patient with severe Legionella pneumonia and dialysis-dependent acute kidney injury persisting for more than 1 month, in whom gallium-67 scintigraphy suggested ongoing renal inflammation in the setting of prolonged tubular injury and supported a diagnostic trial of corticosteroid therapy. Subsequent biopsy showed acute tubular injury with chronic tubulointerstitial changes and a focal, reactive T-cell infiltrate, and renal function was fully restored, suggesting that gallium-67 scintigraphy may be a useful adjunctive finding when biopsy is not immediately feasible.

Case report

A 73-year-old man was transferred to our hospital with severe respiratory failure and multiorgan failure. Sixteen days earlier, he had participated in a 10-km road race and visited a hot spring facility. He subsequently developed malaise, myalgia, and a productive cough. Despite treatment at a local clinic, his symptoms worsened. Three days before admission, he presented to the referring hospital with a fever of 39 °C. Although antibiotic therapy was initiated, he rapidly developed respiratory failure requiring mechanical ventilation.

He arrived at our hospital on Day 1 while sedated and mechanically ventilated. His vital signs were as follows: heart rate, 61 beats/min; blood pressure, 104/48 mmHg; and body temperature, 39.4 °C. Physical examination revealed coarse crackles in both lungs, and no skin rash was present. Chest imaging showed bilateral consolidations, predominantly in the left lung. Admission laboratory findings (Table 1) showed severe systemic inflammation (white blood cell count 24,700/μL; C-reactive protein 50.75 mg/dL) and markedly impaired renal function (serum creatinine 2.94 mg/dL; estimated glomerular filtration rate [eGFR] 17.4 mL/min/1.73 m²; blood urea nitrogen 42.9 mg/dL). Peripheral eosinophilia was absent. At the referring hospital, serum CK had peaked at 10,230 U/L 4 days before transfer and had declined progressively (3033 U/L 2 days before transfer; 671 U/L 1 day before transfer) to 482 U/L by the time of transfer to our hospital, indicating a resolving rather than mild episode of rhabdomyolysis. Urinalysis showed markedly elevated β2-microglobulin (β2-MG) at 9959 μg/L (approximately 33-fold the upper limit of normal [299 μg/L]) together with granular and waxy casts, suggesting extensive tubular injury. Urinary eosinophils were examined and were negative. The urinary Legionella antigen test was positive, and the patient was diagnosed with severe Legionella pneumonia complicated by AKI.

Table 1.

Laboratory data on admission

Biochemistry		Hematology
TP	6.9 g/dL	WBC	24,700/μL
Alb	2.5 g/dL	Neut	95.0%
Cr	2.94 mg/dL	Lymph	1.0%
BUN	42.9 mg/dL	Mono	4.0%
eGFR	17.4 mL/min/1.73 m2	Eosino	0%
UA	6.6 mg/dL	Baso	0%
Na	139 mEq/L	RBC	336×104/μL
K	4.7 mEq/L	Hb	10.9 g/dL
Cl	102 mEq/L	Hct	32.9%
Ca	8.2 mg/dL	PLT	16.9×104/μL
P	5.5 mg/dL	Urinalysis
Mg	2.7 mg/dL	Specific gravity	1.016
T-Bil	1.6 mg/dL	pH	6.0
γ-GTP	67 U/L	Protein	(1 +)
AST	132 U/L	Occult blood	(1 +)
ALT	144 U/L	RBC	50~99/HPF
LDH	476 U/L	U-P/Cr	1.7 g/gCr
CK	482 U/L	Granular cast	3~9/10LPF
Immunology and serology		Waxy cast	5~9/WF
CRP	50.8 mg/dL	U-β2MG	9959.0 μg/L
Legionella antigen (urine)	Positive	NAG	32.2 U/L

Alb, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Baso, basophil; BUN, blood urea nitrogen; CK, creatine kinase; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; Eosino, eosinophil; γ-GTP, γ-glutamyl transpeptidase; Hb, hemoglobin; Hct, hematocrit; HPF, high-power field; LDH, lactate dehydrogenase; LPF, low-power field; Lymph, lymphocyte; Mono, monocyte; NAG, N-acetyl-β-D-glucosaminidase; Neut, neutrophil; PLT, platelet; RBC, red blood cell; T-Bil, total bilirubin; TP, total protein; UA, uric acid; U-β2-MG, urinary β2-microglobulin; U-P/Cr, urinary protein-to-creatinine ratio; WBC, white blood cell; WF, whole field

Empiric intravenous levofloxacin and piperacillin/tazobactam were started. As the respiratory condition gradually improved, piperacillin/tazobactam and levofloxacin were discontinued on Days 7 and 9, respectively. The patient was weaned from mechanical ventilation on Day 16. For renal support, continuous hemodiafiltration (CHDF) was initiated on Day 2 because of progressive azotemia (blood urea nitrogen > 100 mg/dL), congestive heart failure, and persistently rising serum creatinine despite preserved urine output. After hemodynamic stabilization, kidney replacement therapy was transitioned to intermittent hemodialysis (IHD). Although his respiratory status stabilized and CK levels normalized, renal function failed to recover, and he was transferred to our department on Day 28 for further evaluation.

At transfer, urinalysis showed persistent proteinuria (1.71 g/gCr), and urinary β2-MG had further increased to 118,539 μg/L, suggesting severe ongoing tubulointerstitial injury despite discontinuation of antibiotics 3 weeks earlier. The patient also had persistent mild disturbance of consciousness of undetermined cause; lumbar puncture and brain MRI showed no explanatory abnormality. The main clinical question was whether the persistent renal failure represented prolonged but self-limited ATN or whether active TIN was also present and might still respond to corticosteroid therapy. As renal biopsy was considered unsafe in this setting, gallium-67 scintigraphy was performed on Day 31 to characterize the ongoing renal inflammation noninvasively. The scan showed diffuse intense uptake in both kidneys, exceeding that in the liver and vertebral bodies (Fig. 1a). Given the gallium-67 findings, which raised the possibility of corticosteroid-responsive TIN and thus a chance of dialysis withdrawal, prednisolone was started at 30 mg/day (0.5 mg/kg/day), a dose within the standard range used for acute interstitial nephritis [4], as diagnostic therapy (Fig. 2).

Fig. 1.

Gallium-67 scintigraphy (posterior view) and renal biopsy findings. a The scan obtained on Day 31 reveals diffuse, intense uptake of gallium-67 in both kidneys, significantly exceeding the uptake in the liver and vertebral bodies. This finding suggested active renal inflammation. b Low-power light microscopy of the renal biopsy (periodic acid-Schiff stain; original magnification × 100), showing the overall cortical parenchyma from both biopsy cores, including multiple glomeruli that are relatively preserved. c Light microscopy of the renal biopsy (periodic acid-Schiff stain; original magnification × 400), showing representative acute tubular injury with adjacent focal interstitial inflammatory-cell infiltration. Arrows point to representative tubular injuries characterized by loss of brush borders and epithelial cell swelling; this single high-power field does not by itself establish the overall distribution of these findings across the biopsy. d CD3 immunostaining (original magnification × 200) demonstrating that the interstitial infiltrate was predominantly composed of T lymphocytes, with focal areas of clustering; in some tubules, CD3-positive lymphocytes were identified within the tubular epithelium, consistent with tubulitis

Fig. 2.

Clinical course of the patient. The graph illustrates the clinical course, highlighting the changes in serum creatinine (sCr) and urinary β2-MG levels. After prednisolone was initiated on Day 31, sCr and urinary β2-MG declined, and hemodialysis was discontinued on Day 43. Abbreviations: CHDF, continuous hemodiafiltration; IHD, intermittent hemodialysis; LVFX, levofloxacin; PIPC/TAZ, piperacillin/tazobactam; PSL, prednisolone

Corticosteroid therapy was followed by clinical improvement. From Day 37, approximately 1 week after corticosteroid initiation on Day 31, urine output increased from 500 to 2000 mL/day, and serum creatinine (measured immediately before each hemodialysis session throughout the treatment course) began to decline. This temporal pattern is also consistent with the expected oliguric-to-diuretic transition of spontaneously recovering ATN, and we cannot exclude this alternative explanation. The disturbance of consciousness also improved after corticosteroid therapy, although brain MRI and cerebrospinal fluid findings were unremarkable and its mechanism remained uncertain.

After improvement in the disturbance of consciousness, renal biopsy was safely performed on Day 39. Light microscopy showed acute tubular injury with chronic tubulointerstitial changes. Focal lymphocytic infiltration was present in the tubulointerstitium together with diffuse tubular injury (Fig. 1c). Although the biopsy was obtained 8 days after corticosteroid initiation, these inflammatory infiltrates were still evident. Immunofluorescence studies were negative. These findings were consistent with acute tubular injury and chronic tubulointerstitial changes, with a focal, reactive T-cell infiltrate; because biopsy was performed late in the clinical course and after corticosteroid initiation, prior or partially treated TIN could not be excluded, although this remains an inference rather than a diagnostic conclusion. Maintenance dialysis was successfully discontinued on Day 43. Serum creatinine continued to improve, and the prednisolone dose was tapered to 20 mg/day on Day 44. The patient was discharged on Day 49, and at outpatient follow-up on Day 57, renal function had returned to the pre-illness baseline documented in the referring hospital records (serum creatinine 0.75 mg/dL; eGFR 77.6 mL/min/1.73 m²).

Discussion

This case illustrates that renal recovery is possible even after more than 1 month of dialysis dependence in Legionella-associated acute kidney injury (AKI), and highlights the potential adjunctive role of gallium-67 scintigraphy in this setting. Although AKI in severe Legionnaires’ disease is usually attributed to dehydration, shock, or rhabdomyolysis-related acute tubular necrosis (ATN), our patient had renal biopsy findings showing acute tubular injury with chronic tubulointerstitial changes and a focal, reactive T-cell infiltrate, and renal function ultimately recovered after corticosteroid therapy. The main clinical challenge was that biopsy was not feasible early in the course, so the possibility of treatable interstitial inflammation had to be weighed noninvasively. Diffuse intense bilateral renal uptake on gallium-67 scintigraphy, together with markedly elevated urinary β2-MG, provided supportive evidence of ongoing renal inflammation and informed the decision to initiate corticosteroid therapy; the subsequent biopsy, however, did not establish active TIN, and the possibility of spontaneous recovery from ATI/ATN could not be excluded.

Renal biopsy in the present case demonstrated acute tubular injury with chronic tubulointerstitial changes. The biopsy contained 19 glomeruli, none of which showed global sclerosis or crescent formation. The interstitium showed lymphocytic infiltration involving approximately 20% of the cortical area, without eosinophils, together with interstitial fibrosis involving approximately 30% of the cortical area. Masson trichrome staining also demonstrated interstitial edema, evident as pale, loosely textured widening of the interstitium between tubules, distinct from the more densely collagenous areas of fibrosis. Tubular atrophy was present in approximately 40% of the cortical area. We note that epithelial flattening associated with acute tubular injury can be difficult to distinguish from established tubular atrophy on a single biopsy, and the extent of chronic-appearing change reported here may not preclude the full functional recovery that was subsequently observed. Hyaline casts were moderately present; myoglobin casts were not identified. Loss of the tubular brush border was noted in a segmental distribution. CD3 immunostaining confirmed that the infiltrating cells were predominantly T lymphocytes, distributed within the interstitium with focal areas of clustering; in some tubules, CD3-positive lymphocytes were identified within the tubular epithelium, consistent with tubulitis. We acknowledge that spontaneous recovery from ATN cannot be completely excluded. The biopsy findings, including prominent tubular injury, interstitial fibrosis, and tubular atrophy, suggest that acute tubular injury and chronic tubulointerstitial damage were major components of the persistent renal failure. The CD3-positive infiltrate was focal rather than diffuse, and this pattern may represent a reactive response to chronic tubulointerstitial injury rather than an independent, corticosteroid-responsive process. The documented episode of rhabdomyolysis likely contributed substantially to the initial severe acute tubular injury; normalization of CK by admission does not necessarily indicate that established tubular necrosis or cast-related injury had already resolved, and the absence of myoglobin casts in the Day 39 biopsy does not exclude earlier myoglobin-associated injury. Nonetheless, several observations are compatible with a possible contribution from active interstitial inflammation: dialysis dependence persisted for more than 4 weeks, longer than typically expected for uncomplicated ATN after removal of the presumed triggers; urinary β2-MG continued to rise despite antibiotic discontinuation; gallium-67 scintigraphy showed intense diffuse bilateral uptake; and clinical improvement followed corticosteroid initiation, with biopsy performed 8 days later still showing residual lymphocytic infiltration. Given the focal nature of the T-cell infiltrate and the prominence of tubular injury and chronic change on biopsy, we cannot determine with confidence whether active interstitial inflammation, as opposed to prolonged recovery from ATI/ATN with a reactive infiltrate, was the dominant driver of renal recovery in this case.

Drug-induced interstitial nephritis should also be considered in patients exposed to multiple antibiotics, and cannot be excluded solely on the basis that antibiotics had been discontinued several weeks earlier, since drug-induced interstitial nephritis can persist, and may still respond to corticosteroids, even after withdrawal of the offending agent. In this case, piperacillin/tazobactam and levofloxacin were discontinued on Days 7 and 9, respectively, yet renal function continued to deteriorate over the subsequent two weeks, remained dialysis-dependent at Day 28, and urinary beta2-microglobulin continued to rise to 118,539 μg/L, a course longer than typically reported for drug-induced interstitial nephritis, which usually improves within days to weeks after cessation of the offending agent [4]; however, this timing alone should not be taken as strong evidence against drug-induced AIN, and the absence of eosinophilia or urinary eosinophils does not exclude this diagnosis. Given the patient’s exposure to piperacillin/tazobactam and levofloxacin, severe Legionella pneumonia, rhabdomyolysis, and critical illness, the relative contribution of drug-induced AIN, infection-associated interstitial inflammation, and other factors to the observed tubulointerstitial inflammation cannot be determined from the available data. Legionella-associated AKI is itself heterogeneous and may reflect systemic inflammation, sepsis-associated tubular injury, or infection-associated immune mechanisms; Legionella is an intracellular pathogen that replicates within macrophages [5], and ongoing antigen presentation by infected macrophages has been proposed as a mechanism that could sustain T-cell-mediated interstitial inflammation after the systemic infection has begun to improve [6]. This mechanism is speculative rather than the most likely explanation: no renal tissue test in this case demonstrated Legionella organisms, Legionella antigen, Legionella PCR positivity, macrophage infection, or a Legionella-specific T-cell response, and macrophage antigen presentation is itself a general immunological mechanism rather than one specific to Legionella infection.

Our review of the literature identified 13 biopsy-proven cases of Legionella-associated TIN, ATN, and mixed TIN/ATN (Table 2) [7–19]. None of these previously reported cases described the use of gallium-67 scintigraphy; to our knowledge, the present case is the first to document its use in this biopsy-proven cohort. The pathological descriptions in these reports were heterogeneous; however, the diagnostic criteria and the timing of kidney biopsy were not standardized across studies. In the present case, the biopsy showed acute tubular injury with chronic tubulointerstitial changes and a focal, reactive T-cell infiltrate; prior or partially treated TIN could not be excluded. Clinical outcomes are similarly variable. Some patients recover with antimicrobial therapy alone, whereas others have only partial recovery despite corticosteroid treatment. Notably, three of the previously reported dialysis-requiring cases [8, 11, 16] recovered with antibiotic therapy alone, without corticosteroids; this observation does not exclude the possibility that our patient would also have recovered without corticosteroid therapy. However, where the duration was reported, these courses were considerably shorter than that of our patient: in Ref. [8], peritoneal dialysis was discontinued after approximately 3 weeks; in Ref. [11], hemodialysis was performed only on days 3 and 5, with the patient discharged on day 15; and in Ref. [16], four hemodialysis sessions had been administered by day 17, although the exact day of discontinuation was not reported. Our patient’s dialysis-dependent course, exceeding 1 month, was longer than in any of these three antibiotic-only cases for which timing data were available. This heterogeneity suggests that prognosis is determined not simply by the presence of TIN, but by the balance between reversible active inflammation and irreversible structural damage such as tubular loss or interstitial fibrosis at the time treatment is started.

Table 2.

Biopsy-proven renal involvement in Legionella pneumonia: reported cases

No. [References]	Age/sex	Renal histology	Peak sCr (mg/dL)	Rhabdomyolysis (CK, U/L)	Dialysis	Steroid (regimen)	Steroid initiation	Gallium-67 scintigraphy	Outcome
1 [7]	47/M	TIN	10.7	Yes (26,021)	Yes	Yes (PSL 40 mg/day)	Day 23	No	Recovered
2 [8]	44/M	TIN	8.6	No (548)	Yes	No (antibiotics only)	–	No	Recovered
3 [9]	N/A	TIN	N/A	N/A	Yes	Yes (N/A)	N/A	No data	Recovered
4 [10]	43/M	TIN	5.3	No (N/A)	No	No (antibiotics only)	–	No	Recovered
5 [11]	26/M	TIN	12.0	Yes (8529)	Yes	No (antibiotics only)	–	No	Recovered
6 [12]	73/M	TIN	7.03	No (N/A)	Yes	No (antibiotics only)	–	No	Partial recovery
7 [13]	55/M	TIN	11.3	Yes (2000)	Yes	Yes (1 mg/kg/day)	Day 4	No	Recovered
8 [14]	51/M	TIN/ATN	4.72	No (523)	No	Yes (oral MP)	Day 4	No	Partial recovery
9 [15]	54/M	TIN	N/A	Yes (N/A)	Yes	Yes (oral MP)	N/A	No	Recovered
10 [16]	54/M	ATN	3.5	Yes (52,656)	Yes	No (antibiotics only)	–	No	Recovered
11 [17]	51/M	ATN	7.6	No (87)	No	No (antibiotics only)	–	No	Recovered
12 [18]	45/M	ATN	7.66	N/A	Yes	No (antibiotics only)	–	No	N/A
13 [19]	53/M	ATN/Myoglobin casts	9.06	Yes (16,000)	Yes	Yes (high dose) + Rituximab	Day 4	No	Recovered
Our case	73/M	ATN*	6.06	Yes (10,230)	Yes	Yes (PSL 0.5 mg/kg/day)	Day 31	Yes	Recovered

ATN, acute tubular necrosis; CK, creatine kinase; F, female; M, male; MP, methylprednisolone; N/A, not available; PSL, prednisolone; sCr, serum creatinine; TIN, tubulointerstitial nephritis

Rhabdomyolysis defined as serum CK > 1000 U/L

"Recovered": return of sCr to baseline or normal range. "Partial Recovery": improvement but persistent elevation above baseline

^*Biopsy findings also included chronic tubulointerstitial changes and a focal, reactive T-cell infiltrate (see “Discussion”)

The present case adds an important practical point to this literature. Even after a prolonged course requiring dialysis, corticosteroid-responsive disease may still be present if active interstitial inflammation persists. Accordingly, the duration of renal failure alone should not be used to exclude the possibility of treatable TIN. Instead, treatment decisions should be guided by whether there is evidence of ongoing inflammation and by whether irreversible damage is thought to predominate.

The main diagnostic challenge in this case was distinguishing TIN from ATN when biopsy could not be performed safely. Two noninvasive findings were compatible with ongoing tubular injury and possible renal inflammation. First, urinary β2-MG was markedly elevated and continued to rise after withdrawal of antibiotics, indicating persistent tubulointerstitial injury beyond the expected recovery window for uncomplicated ATN, although this marker is not specific for TIN. Urinary NAG was also measured in this patient (range 3.9–56.3 U/L over the course of hospitalization) but, like β2-MG, is not specific for TIN. Second, gallium-67 scintigraphy demonstrated diffuse intense bilateral renal uptake clearly exceeding that of the liver and vertebral bodies, corresponding to grade 5 on the semiquantitative grading scale used in clinical practice, in which uptake exceeding that of the vertebral bodies is considered highly suggestive of active renal inflammation. This pattern has been reported to favor active interstitial nephritis over ATN [2, 3]. Whole-body planar gallium-67 scintigraphy (anterior and posterior projections) combined with SPECT was performed approximately 72 h after tracer injection, in accordance with standard imaging timing for renal evaluation; uptake grade was determined visually according to a previously published grading scale, based on comparison with hepatic and vertebral body uptake. In a retrospective clinical study, intense gallium-67 uptake (grade 5) showed excellent specificity and positive predictive value for acute interstitial nephritis, whereas a cutoff of 3 provided a sensitivity of 61% and a specificity of 75% [3]. This grading scale is supported by a single small retrospective clinical study [3] and an experimental animal model [2], and its diagnostic performance has not been validated in larger cohorts. Region-of-interest analysis of the posterior whole-body image confirmed that renal uptake exceeded hepatic and vertebral body uptake bilaterally (kidney-to-liver ratio 1.27–1.28; kidney-to-vertebral body ratio 1.39–1.40), consistent with the visual grade 5 assessment. Taken together, these findings supported ongoing renal inflammation in the setting of prolonged tubular injury and informed the decision to initiate corticosteroids before biopsy became possible. The subsequent increase in urine output beginning approximately 1 week after corticosteroid initiation and the fall in creatinine, together with biopsy evidence of residual lymphocytic infiltration 8 days after steroid initiation, are compatible with gallium-67 uptake having reflected active renal inflammation in this case; however, as discussed above, the relative contribution of active interstitial inflammation versus a reactive infiltrate accompanying prolonged tubular injury cannot be determined with confidence, and gallium-67 findings alone do not distinguish TIN from ATN with accompanying inflammatory repair, nor do they establish corticosteroid responsiveness. This report also has important limitations: it describes a single patient, biopsy was obtained after corticosteroid initiation and may therefore have underestimated the inflammatory burden, and spontaneous recovery from ATN cannot be excluded. Although serum creatinine was measured consistently before each hemodialysis session, we cannot fully exclude the influence of dialysis dose, ultrafiltration volume, or catabolic state on the observed trend during the period of intermittent hemodialysis; the subsequent increase in urine output, successful discontinuation of dialysis on Day 43, and return of renal function to the pre-illness baseline by Day 57 provide more dialysis-independent evidence of renal recovery. The diagnosis of Legionella infection was based solely on a positive urinary antigen test; sputum culture, PCR, and serogroup determination were not performed, which is an additional limitation. Because kidney biopsy was ultimately performed in this case, gallium-67 scintigraphy is best regarded as ancillary evidence of renal inflammation, potentially most useful when biopsy is not immediately feasible, rather than as a decisive tool for differentiating TIN from ATN. It is not widely available, and its sensitivity and specificity for TIN in the setting of Legionella infection have not been formally established; its interpretation should always be integrated with the overall clinical picture.

In conclusion, this case shows that renal recovery remains possible in Legionella-associated AKI even after more than 1 month of dialysis dependence, and that the duration of renal failure alone should not be used to exclude the possibility of treatable interstitial inflammation. Gallium-67 scintigraphy, interpreted alongside serial tubular injury markers, may provide ancillary noninvasive evidence of ongoing renal inflammation and help inform the decision to initiate corticosteroid therapy when kidney biopsy is not immediately feasible, although it does not by itself establish the diagnosis of TIN or predict corticosteroid responsiveness.

Declarations

Conflict of interest

The authors have declared that no conflict of interest exists.

Ethical approval

This case report was conducted in accordance with the ethical standards of the Institutional Review Board of Akita University Graduate School of Medicine and with the 1964 Helsinki Declaration and its later amendments. The Institutional Review Board of Akita University Graduate School of Medicine reviewed this case report and confirmed that formal ethical approval was not required.

Informed consent

Written informed consent for publication was obtained from the patient.