- Email: [email protected]
The current role of gallium imaging in infection
The Current Role of Gallium Imaging in Infection Christopher J. Palestro The role of gallium imaging in infection has changed considerably during the past several years. Once the mainstay of radionuclide imaging of infection, it has been supplanted to a very great extent by labeled leukocyte imaging. Despite the success of the labeled white-cell technique, gallium still plays an important role in the radionuclide evaluation of infection. It is not possible, for a variety of reasons, to perform white-cell imaging on all patients, and gallium imaging is certainly an acceptable substitute. In certain circumstances, rather than merely being a substitute, gallium is an important complement to leukocyte imaging. This is best illustrated by the patient with a fever of unknown origin (FUO). Although a negative leukocyte study effectively excludes an acute infection, it fails to identify the source of the patient's fever, a not uncommon situation in view of the fact that only approxi-
mately 25% of all FUOs are caused by infection, A complementary gallium study under these circumstances may identify either a chronic infectious process or even a neoplasm, conditions for which whitecell imaging is relatively insensitive. Although leukocyte imaging is probably superior to gallium for most infections of the musculoskeletal system, this technique is of limited value in patients with suspected vertebral osteomyelitis. There are data that suggest that sequential bone gallium imaging may be a better way to diagnose this entity. Finally, in immunocompromised patients, gallium imaging is clearly the procedure of choice for detecting the opportunistic respiratory infections and lymph node abnormalities that are so prevalent in this population.
HE IDEAL T E C H N I Q U E for localizing T infection should be sensitive, specific, easily prepared, and rapidly completed. Although
24 hours after injection as opposed to 48 hours or more for gallium. However, leukocyte imaging is not without its own disadvantages. The technique of labeling is an in vitro process that is labor intensive and requires sterile precautions. In addition to the hazards associated with the handling of (potentially) infected blood products, not all institutions can meet the requirements for, or have access to, facilities capable of labeling white cells. There are patients who for personal or religious reasons may be unwilling to undergo labeled leukocyte imaging, Severely leukopenic patients and those with sickle cell disease (whose red cells do not consistently sediment) may not be suitable candidates for labeled white cell imaging. In most clinical settings, although a mixed population of leukocytes is labeled, the majority of the cells labeled are neutrophils. Consequently, the procedure will be most useful in conditions that incite a neutrophilic response, ie, acute bacterial infection. When the response is other than neutrophilic, or the neutrophilic response has waned, leukocyte imaging is less successful. Under these circumstances, 67Ga imaging can be most rewarding.
neither 67Ga- nor 111In-labeled leukocyte imaging satisfy all these requirements, both techniques enjoy widespread popularity and, when used appropriately, can provide valuable diagnostic information. JHin.leukocyte imaging offers several advantages over gallium and is currently the radionuclide procedure of choice for localizing most types of infection. The advantages of this technique include a virtually constant physiological distribution of radiotracer, which is limited to the liver, spleen, and bone marrow, in contrast to gallium, where the normal distribution of activity can vary considerably from individual to individual. Moreover, the tracer does not accumulate in the normal genitourinary and gastrointestinal tracts, facilitating the detection of infection in these systems. Uptake of labeled leukocytes in normally healing wounds, tumors, and bony trauma occurs less often than with gallium, ie, the study is more specific. Finally, the results of labeled leukocyte studies are routinely available within
From the Division of Nuclear Medicine, Long Island Jewish Medical Center, New Hyde Park, NZ. Address reprint requests to Christopher J. Palestro, MD, Division of Nuclear Medicine, Long Island Jewish Medical Center, Lakeville Road, New Hyde Park, NYl1042. Copyright 9 1994 by W..B. Saunders Company 0001-2998/94/2402-0003505. O0/ 0 128
Copyright 9 1994 by W.B. Saunders Company
87GA
Uptake of 67Ga by tumors was originally reported in 1969, whereas its localization in inflammation was described shortly thereafter. 1,2 From the early 1970s through the mid1980s, 67Ga was the mainstay of radionuclide
Seminars in Nuclear Medicine, Vol XXIV, No 2 (April), 1994: pp 128-141
GALLIUM AND INFECTION
imaging of inflammation and infection (along with 99mTc-methylene diphosphate [MDP] for musculoskeletal infection), and even today among the myriad of techniques currently (or soon to be) available, this tracer is still an important tool in the diagnostic armamentarium of nuclear medicine. Gallium-67 is a cyclotron-generated radiopharmaceutical that can be produced from targets of enriched, stable zinc isotopes by a variety of nuclear reactions using either deuterons or protons. It emits gamma rays over a broad range of 93 keV to 880 keV, with the following four principal energy peaks suitable for imaging: 93 keV, 184 keV, 296 keV, and 388 keV. It has no beta emissions. Gallium-67 is a group III-b transition metal similar to the ferric ion in terms of its charge,
129
atomic radius, and the inorganic complexes it forms. Unlike iron, however, it cannot be reduced in vivo and does not react with protoporphyrin IX to form heme. 3 Approximately 10% to 25 % of the tracer dose is excreted via the kidneys during the first 24 hours after intravenous injection after which time the principal route of excretion is the colon. By 48 hours after injection, approximately 75% of the injected dose remains in the body and is equally distributed among the liver, bone and bone marrow, and soft tissues. 3 This normal distribution can, however, be very variable and may confound image interpretation (Fig 1). The precise mechanisms of 67Ga localization in infection are not thoroughly understood, but it is likely that several factors are involved. Gallium-67 is incorporated into leukocytes, pre-
Fig1. (A)Anteri•re7Gawh••e-b•d•image•h•wsn•rma•physi•••gica•distributi•n•fr•di•tracer•nthe•iver•b•ne•b•nemarr•w• and soft tissues. Minimal bowel activity is present. (B) Anterior 87Ga whole-body image shows a very different, but still essentially normal, distribution of radiotracer in a woman studied I week postpartum. Intense uptake of tracer is seen in both breasts as well in as the eyes and large bowel. Minimally increased activity is evident in the parotid glands.
130
CHRISTOPHER J. PALESTRO
sumab~y bound to the intraceUular lactoferrin abundant in these cells. Leukocytes, of course, migrate to and localize in loci of infection, and it is likely that at least some of the tracer is transported to sites of infection intracellularly in leukocytes. Gallium-67 uptake in infection in patients without circulating leukocytes has been described, however; therefore, intracellular leukocyte transport is not the sole explanation. When leukocytes localize at sites of inflammation, they excrete at least some of their intracellular lactoferrin, which remains localized and bound to macrophages. Gallium, transported either in ionic form or bound to transferrin, may leak through the vascular epithelium at the site of infection where it is then bound to the lactoferrin. Another uptake mechanism may involve direct uptake of gallium by infective organisms themselves, an occurrence that has been shown in vitro. Micro-organisms grown in a low-iron environment produce siderophores that have an extraordinary binding affinity for gallium as well as iron. Because there is very little free iron present in most tissues, it is assumed that pathogenic micro-organisms produce siderophores, and the siderophore-gallium complex is presumably then transported directly into the cell. 3 For localization of infection in adults, approximately 5mCi (185 MBq) of 67Ga is injected as 67Ga-citrate to facilitate solubilization. Although abnormalities can be detected earlier, imaging is most commonly performed at 48 to 72 hours after injection. The value of cathartics to rid the bowel of normal gallium activity is questionable. 4,5 At our institution, they are not routinely used. To obtain diagnostically useful images, a multipeak gamma camera equipped with a suitable medium energy collimator is mandatory. To facilitate a review of the current role of gallium imaging in infection, patients have been arbitraily divided into three broad categories that are based on the indications most often given for requesting the study: (1) those with fever of unknown origin (FUO); (2) those with suspected osteomyelitis; and (3) those with impaired immunocompetence.
occasions and no diagnosis is established after at least 1 week of inpatient evaluation. 6 Although these are well defined, specific criteria, a review of the imaging literature suggests a more liberal use of the term FUO, which includes febrile patients in whom no definitive diagnosis has been established without regard to the duration or degree of fever or length of hospital stay. Only approximately 25% of all FUOs are caused by infection, and for a technique to be useful, it must be specific; this is particularly true for the abdomen, where nearly one third of the infections responsible for FUOs are located. Although rapid whole-body surveys can be performed with both 67Ga and labeled leukocyte imaging, the advantages of the white-cell technique (described above) make it the current radionuclide procedure of choice for imaging the patient with an FUO with an overall accuracy ranging from 85% to 100%. 7-1~ Nevertheless there is a role for gallium imaging in this population. Sfakianakis et al n prospectively compared gallium imaging with leukocyte imaging for diagnosing occult sepsis. Although the overall accuracy of the two techniques (gallium, 90%; labeled leukocyte, 94%) was nearly identical, the investigators found white-cell imaging to be most useful in patients who were symptomatic for less than 2 weeks, whereas gallium imaging performed better in patients who were symptomatic for more than 2 weeks. They concluded that a negative leukocyte study in a patient symptomatic for more than 2 weeks should be followed by gallium scintigraphy (Fig 2). Under certain circumstances, the nonspecificity of gallium imaging actually may be advantageous. Although a negative white-cell study excludes infection as the cause of an FUO, it does not identify the cause of the fever. A gallium scan performed under these circumstances can, when positive, direct the clinician to a fertile site for additional investigation and may even identify an occult neoplasm. For this reason, for those patients with an FUO and a negative white-cell study, regardless of the duration of symptoms, we routinely recommend gallium scintigraphy (Fig 3).
FUO The strict definition of an FUO is an illness of not less than 3 weeks' duration, during which time a patient's fever exceeds 38.3~ on several
OSTEOMYELITIS
Three-phase bone scintigraphy using [99mTC] MDP is the radionuclide procedure of choice
GALLIUM AND INFECTION
131
Fig 2. Indium-Ill-labeled white cell (left) and STGa (right) images performed on a patient with a 1-month history of fever, elevated white count, and left lower-quadrant pain. The white-cell study is normal, whereas a linear focus of activity extends obliquely through the left lower quadrant on the gallium image (arrow). A psoas abscess was surgically drained.
for diagnosing uncomplicated osteomyelitis, with an accuracy exceeding 90%. 12Abnormalities on bone scintigraphy reflect increased bone-mineral turnover, not infection per se, and this technique has been less successful in diagnosing osteomyelitis superimposed on noninfectious conditions that cause increased bone-mineral turnover. Under these conditions, the specificity of the study has ranged from 0% to 78%. 12 Gallium scintigraphy has been used in conjunction with bone scintigraphy in an effort to enhance the specificity of the technique. Lisbona et a113found that sequential bone/gallium imaging was superior to bone scintigraphy alone for distinguishing cellulitis from osteomyelitis, for precise localization of a focus of infection, and for separating active (acute) from inactive (chronic) osteomyelitis. RosenthaU et aP 4 used combined bone/gallium scintigraphy to evaluate suspected osteomyelitis in patients with orthopedic hardware. They found that infection was not likely to be present when the gallium images were negative (regardless of whether the bone images were positive or negative) or when the distribution of activity was similar on both bone and gallium images (congruent images). They also noted that infection was likely to be present when images were spatially congruent and the uptake of gallium was intense or when the distribution of the two radiotracers was different (spatially incongruent images) (Fig 4). Schauwecker et al, 15 as part of a larger study,
evaluated combined bone/gallium scintigraphy for diagnosing complicating osteomyelitis. Although they found that this technique was excellent for excluding osteomyelitis when the study was normal, and for diagnosing osteomyelitis when the intensity of uptake of 67Ga exceeded the uptake of [99mTc]MDP or when the images were spatially incongruent, these criteria were met in only 28% of the patients studied. These investigators concluded that 67Ga images added little diagnostic information not available on bone scintigraphy alone. In this series, 111In-granulocyte imaging, combined with [99mTc]MDP imaging, was superior to bone/ gallium scintigraphy for diagnosing acute osteomyelitis. Merkel et a116 reviewed sequential bone/ gallium imaging performed on 130 patients with painful orthopedic prostheses. Images were interpreted using criteria similar to those of Rosenthall et al, except that Merkel et al added an equivocal category that included studies in which the distribution and intensity of uptake were similar (or congruent) on both bone and gallium images. These investigators reported an overall accuracy of 77% for the sequential technique. Gomez-Luzuriaga et a117 compared sequential bone/gallium imaging with 111In-leukocytes in 40 patients with suspected total hip replacement infection. The accuracy of the bone/ gallium technique in their series was 80% versus 93% for white-cell imaging. Similar results for
132
CHRISTOPHER J. PALESTRO
Fig 3. Anterior chest image from a ["-Tc]HMPAO-labeled leukocyte study (A) performed on a 9-year-old boy with a I-week history of fever was normal (except for splenomegaly). 67GalIJum-67 chest image (B) shows increased activity in the left upper thorax, Gallium-67 coronal SPECT images (C) confirm that the abnormality extends into the left axilla. Lymph node biopsy showed Hodgkin's lymphoma.
combined bone/gallium imaging have been reported by other investigators, with the overall accuracy of the technique ranging between 60% and 80%. ls-2~ Although initial results using labeled leukocyte imaging for diagnosing complicating osteomyelitis were variable, recent data indicate that when leukocyte imaging is combined with marrow imaging, the diagnostic accuracy of the study consistently exceeds 90%, 2123 making this dual-modality technique the procedure of choice for diagnosing complicated osteomyelitis. The results of white-cell imaging are less than
satisfactory, however, for diagnosing vertebral osteomyelitis. We have reported that the technique is neither sensitive nor specific, with an overall accuracy of approximately 66%. 24In contrast, Modic et a125 evaluated sequential bone/ gallium imaging and reported an accuracy of 86% for diagnosing vertebral osteomyelitis, similar to the accuracy they found for magnetic resonance imaging (MRI). Combined bone/gallium imaging recently has been shown to be useful in both tuberculous and nontuberculous infectious spondylitis. 26 Although the data are limited, they do suggest that bone/gallium imaging may
GALLIUM AND INFECTION
133
Fig 4. Technetium-99-MDP (A) and S~Ga(B) forearm images from an adolescent patient with fractures of the left radius and ulna and the right fourth metacarpal. Bone image shows diffusely increased uptake throughout the left ulna and radius as well as the right fourth metacarpal. Gallium uptake is confined to the left ulnar diaphysis (incongruent distribution) and is intense, consistent with osteomyelitis, which was operatively confirmed. There is virtually no gallium uptake in the left radius, which excludes osteomyelitis. The distribution of the two radiotracers in the right fourth metacarpal is spatially congruent, but the intensity of gallium uptake is much less than that of MDP, and the study therefore is negative for osteomyelitis.
be superior to labeled leukocyte imaging for vertebral osteomyelitis; at the very least, this technique clearly merits further investigation. THE IMMUNOCOMPROMISED PATIENT
The immunocompromised individual has an increased susceptibility to numerous infections that often are caused by organisms not typically pathogenic in the immunocompetent host. These infections can involve virtually every organ system in the body, and despite the often serious nature of these ailments, patients frequently present with minimal signs and symptoms. The ability to perform whole-body surveys rapidly and in a timely fashion, combined with reasonable accuracy, makes radionuclide techniques a particularly attractive diagnostic tool in this population. The discussion that follows focuses on radionuclide imaging of the chest and abdomen, and although based on results obtained in acquired immunodeficiency disease (AIDS) patients, it is applicable to the general population
of immunocompromised people, including those who are receiving antineoplastic therapy and the ever increasing number of organ-transplant recipients. The Thorax
Radionuclide techniques have not figured prominently in the diagnosis of respiratory infection in the general population, probably be, cause sputum cultures and chest radiographs are easily obtainable and usually diagnostic, In contrast to their limited role in the general population, radionuclide studies, notably gallium, have come to play a crucial role in the diagnosis of AIDS-related respiratory disease. Four distinct patterns of gallium uptake in the thorax have been observed, and each pattern is associated with a particular entity or group of entities. The four patterns are (1) normal, (2) lymph-node uptake, (3) focal pulmonary parenchymal uptake, and (4) diffuse pulmonary parenchymal uptake.
134
Normal The normal gallium scan of the chest, particularly in the setting 0 f a normal chest x-ray, excludes infection with a high degree of certainty. Barron et a127 reported that only 1 of 15 patients with negative gallium scans had pulmonary infection, making the predictive value of a negative study 93% in the series. Their 1 false-negative study occurred in a patient with Pneumocystis carinii pneumonia (PCP) who had received 2 weeks of treatment before imaging. None of the eleven patients whose gallium scans and chest x-rays were both negative had pulmonary disease (negative predictive value, 100%). Kramer et a128 reviewed 86 gallium scans performed on 71 human immunodeficiency virus (HIV)-positive individuals and found no evidence of thoracic infection in any of the 20 patients whose gallium studies and chest radiographs both were negative. Woolfenden et a129reported a negative predictive value of 91% when only pulmonary activity greater than marrow activity was the criterion for a positive study. Although the combination of a negative gallium scan and a negative chest x-ray rule out any pulmonary disease with a high degree of certainty, a negative gallium scan in the setting of an abnormal chest x-ray has a different, although no less important, connotation. Among the neoplasms associated with AIDS is Kaposi's sarcoma, which is not gallium avid. In patients with pulmonary Kaposi's sarcoma, a positive gallium scan suggests the presence of superimposed inflammation or infection. A negative scan with a positive chest x-ray indicates that Kaposi's sarcoma is the most likely cause of the patient's deteriorating respiratory status, z8,29 Despite providing strong evidence against pulmonary infection, the negative gallium image occasionally has been observed in patients with clinically obvious progression of respiratory infection, and in this setting, the negative results portend a grave prognosis. Bitran et al 3~ described four such patients, all of whom died within several weeks after undergoing gallium imaging. At autopsy, all were found to have multiple opportunistic infections, including cytomegalovirus, PCP, and Mycobacterium avium intracellulare (MAI). The inflammatory response in each of the patients was abnormal and was characterized by a lack of granulocytes. The
CHRISTOPHER J. PALESTRO
investigators speculated that the absence of galIium Uptake in the Setting Of disseminated infection may have been caused by the aberrant inflammatory response. Lymph node uptake Lymph node uptake of gallium is a frequent finding in the HIV-positive patient (Fig 5). Although such uptake can be identified virtually anywhere in the body, it is most often encountered in the hilar, mediastinal, and para-aortic lymph nodes. Uptake in supraclavicular and cervical lymph nodes often can be seen on gallium images of the chest. Kramer et a128 identified nodal uptake in 10 (12%) of 86 scans performed on HIV-positive patients; this was associated with MAI in 9 patients and with lymphoma in 1 patient. In a later series, 31these same investigators identified nodal uptake in 58 (24%) of 237 scans. MAI or M tuberculosis was present in 35 of the 58 patients, and lymphoma was present in 7. Other conditions associated with nodal uptake of gallium included cryptococcal infection, lymphadenitis, and herpes simplex. Eight (14%) of the 58 patients with nodal uptake had PCP, and all had increased pulmonary parenchymal gallium uptake in addition to nodal uptake. Of particular importance is the fact that in this series, nodal uptake alone was never associated with PCP.
Fig 5. Anterior gallium chest image shows mediastinel, left hilar, and left supraclavicular lymph node uptake as well as faint focal uptake in the right upper-lung parenchyma (arrow) in a patient with MAI pneumonia.
GALLIUM AND INFECTION
Nodal uptake of gallium in the generalized lymphadenopathy associated with the HIVpositive population is variable. Bitran et aP ~ identified lymph node uptake of gallium in 13 (41%) of 32 HIV-positive patients with generalized lymphadenopathy and AIDS-related complex (ARC). Other investigators, however, have not found this entity to be particularly gallium avid. Kramer et aP 2 reported that in their experience, the generalized lymphadenopathy found in patients with ARC is not gallium avid. More recently, Podzamczer et aP 3 reviewed lymph node uptake of gallium in 30 HIVpositive patients with lymphadenopathy. Tissue diagnoses were available in all 30 patients. Lymph node uptake at least as intense as liver uptake was present in 17 patients and was associated with M tuberculosis in 11 patients, MAI in 1 patient, lymphoma in 3 patients, Kaposi's sarcoma with Castelman's disease in 1 patient, and follicular hyperplasia in 1 patient. Three patients had nodal uptake less intense than hepatic uptake, and all had follicular hyperplasia. Ten patients had no gallium uptake in lymph nodes, 9 had follicular hyperplasia, and 1 had Kaposi's sarcoma. Localized (focal) pulmonary uptake This pattern of gallium uptake most frequently is found in the setting of bacterial pneumonia (Fig 6). When the uptake is multifocal, aggressive infections such as actinomycosis or nocardia should be considered. Although gallium studies are often positive in AIDS patients with bacterial pneumonia, corresponding radiographic abnor-
Fig 6. Anterior and posterior eTGa images show focal parenchymal uptake in the right lower lung in a patient with S aureus pneumonia. Note the intense gastric uptake (arrows) in this patient without evidence of gastrointestinal disease,
135
malities invariably are present, and some investigators think that the gallium findings contrib ute little to the diagnosis. 31,34 It is interesting and important to note that other than bacterial pneumonia, the entity most frequently associated with localized pulmonary uptake of gallium is PCP. 31 Diffuse pulmonary uptake. In an effort to maintain consistency in interpretation, facilitate comparison of serial studies, and enhance the diagnostic accuracy of the study, a grading system for evaluating the intensity of pulmonary parenchymal uptake of gallium has evolved. Although the particulars of any given grading system vary somewhat from institution to institution, the intensity of pulmonary gallium uptake commonly is divided into one of four or five grades based on a comparison with some reference standard, which typically includes surrounding soft tissues and the liver. The grading system employed at our institution is as follows (Fig 7): grade 0 (normal), pulmonary uptake is indistinguishable from surrounding soft tissues; grade 1 (equivocal), pulmonary uptake is more intense than grade 0 but less intense than rib activity and is of questionable significance; grade 2 (abnormal), pulmonary uptake is clearly increased but less intense than hepatic activity; a photopenic cardiac silhouette is present; grade 3 (abnormal), pulmonary uptake is increased and equal to hepatic uptake; a photopenic cardiac silhouette is present; and grade 4 (abnormal), pulmonary uptake exceeds liver uptake; a photopenic cardiac silhouette is present.
136
CHRISTOPHER J. PALESTRO
+ i
Fig 7. (A) Grade 0 (normal) gallium image of the chest; (B) Grade 2 diffuse pulmonary uptake; (C) Grade 3 diffuse pulmonary uptake; (D) Grade 4 diffuse pulmonaryuptake {see text for description of grading system).
Diffuse pulmonary uptake of gallium in the AIDS patient is most often associated with PCP, and the overall sensitivity of the technique has generally been reported to exceed 90%. 27-30 Unfortunately, there is a rather long differential diagnosis that accompanies the finding of diffuse pulmonary uptake of gallium, and when this is the sole criterion used for diagnosing PCP, the specificity of the study can be quite low, having been reported to range between 20% and 82%. 31,35 The specificity of the study can be enhanced
by applying additional criteria to the interpretation of the study. Barron et a127 found that a positive gallium scan was more specific for PCP when the chest x-ray was negative than when the x-ray was positive (85% v 74%). Intense pulmonary uptake on gallium images is more likely to be associated with PCP than with other conditions in the AIDS population. Coleman et a135 found that the specificity of the study improved from 20%, when any pulmonary uptake was considered positive for PCP, to 90%, when only pulmonary uptake at least as intense
GALLIUM AND INFECTION
as liver uptake was considered positive. Kramer et a131reported that the specificity of the study increased from 79%, when abnormal lung activity more intense than bone marrow activity was considered positive for PCP, to 95%, when only pulmonary activity more intense than liver activity was considered positive. These investigators also observed that consideration of the distribution of gallium uptake could improve the specificity of the scan for PCP. When diffuse lung uptake was considered positive for PCP, the positive predictive value was 72%; the positive predictive value increasedto 87% when heterogeneous diffuse uptake was the diagnostic criterion for a positive study. Treatment can affect both the intensity and pattern of pulmonary uptake of gallium in AIDS patients. Woolfenden et a129 noted that only 1 of 12 patients with PCP who showed pulmonary uptake at least as intense as liver uptake was being treated at the time of imaging. In contrast, 4 of 7 patients with PCP whose gallium studies showed lung activity less intense than that of the liver were being treated at the time of imaging. Aerosolized pentamidine frequently is used for prophylaxis of recurrent PCP. Gallium uptake, confined to the upper lungs, has been observed in a patient who was treated prophylactically with aerosolized pentamidine and who developed recurrent PCP that was confined to the upper lungs) 6 The proposed mechanism for the regional recurrence of this disease is decreased aerosol deposition in the upper-lung fields. Finally, there are some data that suggest that gallium imaging may be useful to monitor patient response to therapy for PCP. 37,38Touazon et a137 performed gallium imaging on eight patients with PCP subsequently confirmed by fiberoptic bronchoscopy. All eight patients had abnormal gallium scans, including seven patients who had diffuse lung uptake of radiotracer and one patient who had localized uptake. Repeat studies were obtained after treatment in six of the eight patients and were positive in four, three of whom had bronchial washings persistently positive for PCP. The fourth patient subsequently received diagnosing of Coccidioides immitus infection. Both patients with negative follow-up gallium scans had negative bronchial washings.
137
In summary, gallium imaging plays an important role in the diagnosis of AIDS-related respiratory illnesses. A negative scan is strong evidence against an infectious process. Lymphnode uptake most often is associated with mycobacterial disease or lymphoma, whereas focal lung parenchymal uptake usually is seen in the setting of bacterial infection. Although it may be because of a myriad of causes, diffuse lung uptake should be considered suspicious for PCP, especially when the uptake is intense or heterogeneous (Fig 8). The Abdomen The gastrointestinal (GI) tract is the largest lymphoid organ in the human body, and consequently, it is a potential reservoir for the HIV. The GI tract is an important site for HIVinduced immunodeficiency and is predisposed to a broad spectrum of protozoal, fungal, viral, and bacterial diseases that result in considerable morbidity and a not insignificant mortality. 39 Infectious agents that frequently involve the upper GI tract include Candida albicans, herpes simplex virus, cytomegalovirus, and Mycobacterium avium intracellulare. Those organisms with a predilection for the bowel include cryptosporidium, microsporidium, isospora belli, salmonella, clostridium difficile, and herpes simplex. Clinically, patients with AIDS-related illnesses of the upper GI tract, especially the esophagus, present with dysphagia, whereas those with intestinal disorders usually present with abdominal pain, cramping, and diarrhea. Despite the fact that nearly 90% of AIDS patients are affected by HIV-related GI disease at some time during the course of their illness, radionuclide imaging has played a limited role in the diagnostic evaluation of these entities. Whereas there are undoubtedly many reasons for this, two factors are particularly important. Successful treatment of AIDS-related GI disease invariably requires precise identification of the causative agent. In contrast to the thorax, where patterns and intensity of radiogallium uptake often can correctly characterize the identity of a pathogen, no such correlates exist in the abdomen. Moreover, normal excretion of this radiotracer via the large bowel further complicates matters by making differentiation of normal from abnormal bowel activity difficult (Fig 9).
138
CHRISTOPHER J. PALESTRO
'TGa Image
I
I
I
I
I
Normal
Focal Nodal Uptake
Focal Parenchymal Uptake I
Diffuse Parenchymal Uptake
I CXR(-)
No pulmonary disease
I CXR(+)
Kaposi's sarcoma
I
I
I
M. tuberculosis MAI
Bacterial pneumonia
Mild Homogeneous
Intense I
Lymphoma
PCP
PCP TreatedPCP CMV MAI Interstitial pneumonitis
I
I CXR(-)
I Heterogeneous
1 PCP
Fig 8. Outline of patterns of thoracic uptake of eTGa and the entities most often associated with these patterns in AIDS patients.
Fig 9. Intense colonic gallium activity in two AIDS patients. (A) Image is from a patient with pseudomembranous colitis. (B) Image is from a person with no gastrointestinal disease.
GALLIUM AND INFECTION
Thete are scant data available pertaining to gallium scintigraphy of HIV-related illnesses !nvolving the esophagus and stomach. Although a case of gallium uptake in esophageal candidiasis has been observed, the investigators reviewed 100 additional gallium scans, performed primarily on immunocompromised patients, and did not detect any additional cases, noting that a major limitation to gallium imaging of this entity is the presence of sternal and vertebral activity. 4~ Gallium accumulation in gastric lymphoma of an AIDS patient has been described41; yet, gastric uptake, in the absence of any pathology, also has been reported in this population, and the presence of gastric activity therefore must be interpreted with caution (Fig 6). 28,42 In the intestinal tract itself, although gallium accumulation in HIV-related small bowel disease has been reported, 32 the vast majority of the data available pertain to gallium scintigraphy of HIV colitides. Woo!fenden et a129 reviewed large-bowel uptake of radiogallium in 95 AIDS patients (164 scans). Images were performed 48 to 72 hours after injection of 3.5 mCi (130 MBq) (3.5 mCi) of 67Ga. Cathartics were not routinely used. The intensity of colonic uptake was graded on a scale of 0 to 5, with 0 being indistinguishable from background and 5 greater being than liver. Thirty-six patients (38%) showed colonic activity more intense than hepatic uptake (grade 5) in at least one segment. Twenty-three of the 36 patients (64%) were symptomatic with either abdominal pain or diarrhea or both. Fourteen of the 36 patients had colonoscopy or microbiological studies. Eight of the 14 (57%) were diagnosed with GI infection. No definite evidence of infection could be found in the remaining 6 patients, all of whom had normal endosc0pies. Three of these 6 individuals underwent colonic biopsy, and all had evidence of chronic inflammation, Of the 59 patients with bowel uptake less intense than hepatic activity, 4 (7%) had GI infections. These investigators concluded that although the finding of colonic activity on gallium images should be interpreted cautiously, unusually intense (large bowel) activity should be considered suspicious for infection. Tatsch et a143 evaluated the reliability of gallium scintigraphy for detection of AIDSrelated intestinal infections in 25 HIV-positive patients. Imaging was performed at 48 and 72
139
hours after injection of 5 mCi (185 MBq) of 67Ga, and cathartics were routinely used before imaging. Studies were considered abnormal only when bowel activity was unchanged between 48 and 72 hours images (regardless of intensity). Seventeen of the 25 patients (68%) had abnormal gallium accumulation; in 14 the abnormalities involved either all or part of the colon, and in the remaining 3 the abnormalities were localized to abdominal lymph nodes. All 12 of the 17 patients with abnormal gallium uptake who underwent diagnostic evaluation had confirmed AIDS-related diseases; 10 patients with abnormal colonic activity had GI infection, and 2 patients with para-aortic lymph node uptake had atypical mycobacterial infection (Fig 10). The 5 remaining patients all had abnormal colonic activity; 2 patients had diarrhea and negative stool cultures and underwent no further GI workup. Three patients were totally asymptomatic and underwent no GI workup whatsoever. No information was provided about the 8 patients with negative abdominal gallium studies. Slizofski et a144 retrospectively reviewed 21 67Ga studies performed on 19 AIDS patients. Patients received 10 mCi (370 MBq) of 67Ga and were usually imaged at 48 to 72 hours after injection. The investigators did not indicate whether or not cathartics were used. Fourteen of the 19 patients underwent more than one set of images. Colonic activity was classified as abnormal when it was more intense than hepatic activity and did not move on successive scans. Seven patients had acute GI disease; gallium scans were true positive in 4 patients and false negative in 3 patients. Among the 14 studies performed on patients without acute GI disease, there were 10 true negatives and 4 false positives. Not surprisingly, the results of gallium imaging in the abdomen are more variable than those in the chest. Although absence of any colonic activity probably excludes significant bowel pathology, the presence of such activity, even when very intense, must be interpreted cautiously. There are few data available comparing gallium with labeled leukocyte imaging in the immunocompromised population. Based on a comparison of these two techniques in 36 febrile AIDS patients, Fineman et a145 concluded that
140
CHRISTOPHER J. PALESTRO
O Fig 10. (A) 67Ga uptake in celiac and para-aortic lymph nodes in an AIDS patient with MAI. (B) Extensive, intense lymph-node uptake of eTGa in celiac, paraaortie, iliac, inguinal and femoral lymph nodes in an HIV-positive patient with lymphoma.
although labeled leukocyte imaging was superior to gallium imaging for the detection of sinusitis, bacterial pneumonia, and bowel infection, gallium imaging was the procedure of choice for detecting lymph node abnormalities and the opportunistic respiratory infections so prevalent in this population. The role of gallium imaging in infection has
changed dramatically over the past decade, not only because of the development of newer radionuclide techniques, but also because of the ever-increasing number of patients with altered immunocompetence. Rather than resulting in the elimination of gallium imaging for infection, these events have simply redefined its indications.
REFERENCES 1. Edwards CL, Hayes RL: Tumor scanning with 67Ga citrate. J Nucl Med 10:103-105, 1969 2. Lavender JP, Lowe J, Barker JR, et al: Gallium-67 citrate scanning in neoplastic and inflammatory lesions. Br J Radio144:361-366, 1971 3. Hoffer P: Gallium: Mechanisms. J Nucl Med 21:282285, 1980 4. Zeman RK, Ryerson TW: The value of bowel preparation in Ga-67 citrate scanning: concise communication. J Nucl Med 18:886-889, 1977 5. Silberstein EB, Fernandez-Ulloa M, Hall J: Are oral cathartics of value in optimizing the gallium scan? Concise communication. J Nucl Med 22:424-427, 1981 6. Petersdorf RG, Beeson PB: Fever of unexplained origin: Report on 100 cases. Medicine 40:1-30, 1961 7. McDougall IR, Baumert JE, Lantieri RL: Evaluation of 111In leukocyte whole body scanning. AJR 133:849-854, 1979
8. Schmidt KG, Rasmussen JW, Sorensen PG, et al: Indium-lll-granulocyte scintigraphy in the evaluation of patients with fever of undetermined origin. Scand J Infect Dis 19:339-345, 1987 9. Syrjala MT, Valtonen V, Liewendahl K, et al: Diagnostic significance of indium-Ill granul0cyte scintigraphy in febrile patients. J Nucl Med 28:155-160, 1987 10. Kelly MJ, KalffV, Hicks RJ, et al: 11qn-oxine labelled leukocyte scintigraphy in the detection and localization of active inflammation and sepsis. Med J Aust 152:352-357, 1990 11. Sfakinakis GN, AI-Sheikh W, Heal A, et al: Comparisons of scintigraphy with In-111 leukocytes and Ga-67 in the diagnosis of occult sepsis. J Nucl Med 23:618-626, 1982 12. Schauwecker DS: The scintigraphic diagnosis of osteomyelitis. AJR 158:9-18, 1992 13. Lisbona R, Rosenthall L: Observations on the sequential use of 99mTc-phosphate complex and 67Ga imaging in
GALLIUM AND INFECTION
osteomyelitis, cellulitis, and septic arthritis. Radiology 123: 123-129, 1977 14. Rosenthall L, Lisbona R, Hernandez M, et al: 99mTcPP and 67Ga imaging following insertion of orthopedic devices. Radiology 133:717-721, 1979 15. Schauwecker DS, Park HM. Mock BH. et al: Evaluation of complicating osteomyelitis with Tc-99m MDP, In111 granulocytes, and Ga-67 citrate. J Nucl Med 25:849-853, 1984 16. Merkel KD, Brown ML. Fitzgerald RH Jr: Sequential technetium-99m HMDP-gallium-67 citrate imaging for the evaluation of infection in the painful prosthesis. J Nucl Med 27:1413-1417. 1986 17. Gomez-Luzuriaga MA, Galan V, Villar JM: Scintigraphy with Tc. Ga and In in painful total hip prostheses. Int Orthop 12:163-167. 1988 18. Merkel KD, Brown ML, Dewanjee MK, el al: Comparison of indium-labeled-leukocyte imaging with sequential technetium-gallium scanning in the diagnosis of lowgrade musculoskeletal sepsis. J Bone Joint Surg 67:465-476. 1985 19. Merkel KD, Fitzgerald RH Jr. Brown ML: Scintigraphic examination of total hip arthroplasty: Comparison of indium with technetium-gallium in the loose and infected canine arthroplasty, in Welch RB. (ed): The Hip. Proceedings of the Twelfth Open Scientific Meeting of the Hip Society, Atlanta, GA, 1984, pp 163-192 20. AI-Sheikh W. Sfakianakis GN, Mnaymneh W, et al: Subacute and chronic bone infections: Diagnosis using I n - l l l , Ga-67 and Tc-99m MDP bone scintigraphy, and radiography. Radiology 155:501-506, 1985 21. Palestro CJ, Kim CK, Swyer AJ. et al: Total hip arthroplasty: Periprosthetic 111In labeled leukocyte activity and complementary 99mTc sulfur colloid imaging in suspected infection. J Nucl Med 31:1950-1955, 1990 22. Palestro CJ, Swyer AJ, Kim CK, Goldsmith SJ: Infected knee prosthesis: Diagnosis with In-Ill-leukocyte, Tc-99m sulfur colloid, and Tc-99m MDP imaging. Radiology 179:645-648, 1991 23. Palestro CJ, Roumanas P. Swyer AJ, et al: Diagnosis of musculoskeletal infection using combined In-111 labeled leukocyte and Tc-99m SC marrow imaging. Clin Nucl Med 17:269-273, 1992 24. Palestro CJ, Kim CK, Swyer AJ, et al: Radionuclide diagnosis of vertebral osteomyelitis: Indium-Ill-leukocyte and technetium-99m-methylene diphosphonate bone scintigraphy. J Nucl Med 32:1861-1865, 1991 25. Modic MT, Feiglin DH, Piraino DW, et al: Vertebral osteomyelitis: Assessment using MR. Radiology 157:157166, 1985 26. Lisbona R, Derbekyan V, Novales-Diaz J, et al: Gallium-67 scintigraphy in tuberculous and nontuberculous infectious spondylitis. J Nucl Med 34:853-859, 1993 27. Barron TF, Birnbaum NS, Shane LB, et al: Pneumocystis carinii pneumonia studied by gaUium-67 scanning. Radiology 154:791-793, 1985 28. Kramer EL, Sanger J J, Garay SM, et al: Gallium-67 scans of the chest in patients with acquired immunodeficiency syndrome. J Nucl Med 28:1107-1114, 1987 29. Woolfenden JM, Carrasquil!o JA, Larson SM, et al: Acquired immunodeficiency syndrome: Ga-67 citrate imaging. Radiology 162:383-387, 1987
141
30. Bitran J, Bekerman C, Weinstein R, et al: Patterns of gallium-67 scintigraphy in patients with Acquired Immunodeficiency Syndrome and the AIDS related complex. J Nucl Med 28:1103-1106, 1987 31. Kramer EL, Sanger JH, Garay SM, et al: Diagnostic implications of Ga-67 chest scan patterns in human immunodeficiency virus-seropositive patients. Radiology 170:671676, 1989 32. Kramer EL, Sanger J J: Nuclear Medicine in the management of the AIDS patient, in Freeman LM (ed): Nuclear Medicine Annual, New York. NY, Raven. pp 37-57, 1990 33. Podzamczer D. Ricart I, Bolao F, et al: Gallium-67 scan for distinguishing follicular hyperplasia from other AIDS-associated diseases in lymph nodes. AIDS 4:683-685. 1990 34. Kramer EL. Sanger JJ: Detection of thoracic infections by nuclear medicine techniques in the Acquired Immunodeficiency Syndrome. Radiol Clin North Am 27: 1067-1076. 1989 35. Coleman DL, Hattner RS. Luce JM, et al: Correlation between gallium lung scans and fiberoptic bronchoscopy in patients with suspected pneumocystis carinii pneumonia, and the acquired immunodeficiency syndrome. Am Rev Respir Dis 130:1166-1169, 1984 36. q-upler RH, Turbiner EH: Scintigraphic distribution of Pneumocystis carinii pneumonia in AIDS treated with prophylactic inhaled pentamidine. Clin Nucl Med 16:772. 1991 37. Tuazon CU, Delaney MD. Simon GL, et al: Utility of gallium67scintigraphy and bronchial washings in the diagnosis and treatment of Pneumocystis carinli pneumonia in patients with the acquired immune deficiency syndrome. Am Rev Respir Dis 132:1087-1092, 1985 38. Moser E, Tatsch K, Kirsch C-M, et al: Value of 67Gallium scintigraphy in primary diagnosis and follow-up of opportunistic pneumonia in patients with AIDS. Lung 168: 692-703, 1990 (suppl) 39. Smith PD, Quinn TC, Strober W, et al: Gastrointestinal infections in AIDS (NIH Conference). Ann Int Med 116:63-77, 1992 40. Rundback JH, Goldfarb CR, Ongseng F: Gallium-67 imaging in candidal esophagitis. Clin Nucl Med 15:38-39, 1990 41. Gianfelice D, Rosenthall L, Falutz J: Gallium-67 detection of occult gastric lymphoma in AIDS. A JR 149:305306, 1987 42. McNeill JA, Llaurado JG: Innocent intramural gastric uptake of gallium-67 in a case of AIDS. Clin Nucl Med 1!:123-126, 1986 43. Tatsch K, Knesewitsch P, Matuschke A, et al: 67Gascintigraphy for evaluation of AIDS-related intestinal infections. Nucl Med Comm 11:649-655, 1990 44. Slizofski WJ, Brown SJ, Dadparvar S, et al: Diagnostic significance of apparent abnormal bowel activity on Ga-67 scans in AIDS patients. Clin Nucl Med 16:473-477, 1991 45. Fineman DS, Palestro CJ, Kim CK, et al: Detection of abnormalities in febrile AID S patients with In-Illlabeled leukocyte and Ga-67 scintigraphy. Radiology 170: 677-680, 1989
mately 25% of all FUOs are caused by infection, A complementary gallium study under these circumstances may identify either a chronic infectious process or even a neoplasm, conditions for which whitecell imaging is relatively insensitive. Although leukocyte imaging is probably superior to gallium for most infections of the musculoskeletal system, this technique is of limited value in patients with suspected vertebral osteomyelitis. There are data that suggest that sequential bone gallium imaging may be a better way to diagnose this entity. Finally, in immunocompromised patients, gallium imaging is clearly the procedure of choice for detecting the opportunistic respiratory infections and lymph node abnormalities that are so prevalent in this population.
HE IDEAL T E C H N I Q U E for localizing T infection should be sensitive, specific, easily prepared, and rapidly completed. Although
24 hours after injection as opposed to 48 hours or more for gallium. However, leukocyte imaging is not without its own disadvantages. The technique of labeling is an in vitro process that is labor intensive and requires sterile precautions. In addition to the hazards associated with the handling of (potentially) infected blood products, not all institutions can meet the requirements for, or have access to, facilities capable of labeling white cells. There are patients who for personal or religious reasons may be unwilling to undergo labeled leukocyte imaging, Severely leukopenic patients and those with sickle cell disease (whose red cells do not consistently sediment) may not be suitable candidates for labeled white cell imaging. In most clinical settings, although a mixed population of leukocytes is labeled, the majority of the cells labeled are neutrophils. Consequently, the procedure will be most useful in conditions that incite a neutrophilic response, ie, acute bacterial infection. When the response is other than neutrophilic, or the neutrophilic response has waned, leukocyte imaging is less successful. Under these circumstances, 67Ga imaging can be most rewarding.
neither 67Ga- nor 111In-labeled leukocyte imaging satisfy all these requirements, both techniques enjoy widespread popularity and, when used appropriately, can provide valuable diagnostic information. JHin.leukocyte imaging offers several advantages over gallium and is currently the radionuclide procedure of choice for localizing most types of infection. The advantages of this technique include a virtually constant physiological distribution of radiotracer, which is limited to the liver, spleen, and bone marrow, in contrast to gallium, where the normal distribution of activity can vary considerably from individual to individual. Moreover, the tracer does not accumulate in the normal genitourinary and gastrointestinal tracts, facilitating the detection of infection in these systems. Uptake of labeled leukocytes in normally healing wounds, tumors, and bony trauma occurs less often than with gallium, ie, the study is more specific. Finally, the results of labeled leukocyte studies are routinely available within
From the Division of Nuclear Medicine, Long Island Jewish Medical Center, New Hyde Park, NZ. Address reprint requests to Christopher J. Palestro, MD, Division of Nuclear Medicine, Long Island Jewish Medical Center, Lakeville Road, New Hyde Park, NYl1042. Copyright 9 1994 by W..B. Saunders Company 0001-2998/94/2402-0003505. O0/ 0 128
Copyright 9 1994 by W.B. Saunders Company
87GA
Uptake of 67Ga by tumors was originally reported in 1969, whereas its localization in inflammation was described shortly thereafter. 1,2 From the early 1970s through the mid1980s, 67Ga was the mainstay of radionuclide
Seminars in Nuclear Medicine, Vol XXIV, No 2 (April), 1994: pp 128-141
GALLIUM AND INFECTION
imaging of inflammation and infection (along with 99mTc-methylene diphosphate [MDP] for musculoskeletal infection), and even today among the myriad of techniques currently (or soon to be) available, this tracer is still an important tool in the diagnostic armamentarium of nuclear medicine. Gallium-67 is a cyclotron-generated radiopharmaceutical that can be produced from targets of enriched, stable zinc isotopes by a variety of nuclear reactions using either deuterons or protons. It emits gamma rays over a broad range of 93 keV to 880 keV, with the following four principal energy peaks suitable for imaging: 93 keV, 184 keV, 296 keV, and 388 keV. It has no beta emissions. Gallium-67 is a group III-b transition metal similar to the ferric ion in terms of its charge,
129
atomic radius, and the inorganic complexes it forms. Unlike iron, however, it cannot be reduced in vivo and does not react with protoporphyrin IX to form heme. 3 Approximately 10% to 25 % of the tracer dose is excreted via the kidneys during the first 24 hours after intravenous injection after which time the principal route of excretion is the colon. By 48 hours after injection, approximately 75% of the injected dose remains in the body and is equally distributed among the liver, bone and bone marrow, and soft tissues. 3 This normal distribution can, however, be very variable and may confound image interpretation (Fig 1). The precise mechanisms of 67Ga localization in infection are not thoroughly understood, but it is likely that several factors are involved. Gallium-67 is incorporated into leukocytes, pre-
Fig1. (A)Anteri•re7Gawh••e-b•d•image•h•wsn•rma•physi•••gica•distributi•n•fr•di•tracer•nthe•iver•b•ne•b•nemarr•w• and soft tissues. Minimal bowel activity is present. (B) Anterior 87Ga whole-body image shows a very different, but still essentially normal, distribution of radiotracer in a woman studied I week postpartum. Intense uptake of tracer is seen in both breasts as well in as the eyes and large bowel. Minimally increased activity is evident in the parotid glands.
130
CHRISTOPHER J. PALESTRO
sumab~y bound to the intraceUular lactoferrin abundant in these cells. Leukocytes, of course, migrate to and localize in loci of infection, and it is likely that at least some of the tracer is transported to sites of infection intracellularly in leukocytes. Gallium-67 uptake in infection in patients without circulating leukocytes has been described, however; therefore, intracellular leukocyte transport is not the sole explanation. When leukocytes localize at sites of inflammation, they excrete at least some of their intracellular lactoferrin, which remains localized and bound to macrophages. Gallium, transported either in ionic form or bound to transferrin, may leak through the vascular epithelium at the site of infection where it is then bound to the lactoferrin. Another uptake mechanism may involve direct uptake of gallium by infective organisms themselves, an occurrence that has been shown in vitro. Micro-organisms grown in a low-iron environment produce siderophores that have an extraordinary binding affinity for gallium as well as iron. Because there is very little free iron present in most tissues, it is assumed that pathogenic micro-organisms produce siderophores, and the siderophore-gallium complex is presumably then transported directly into the cell. 3 For localization of infection in adults, approximately 5mCi (185 MBq) of 67Ga is injected as 67Ga-citrate to facilitate solubilization. Although abnormalities can be detected earlier, imaging is most commonly performed at 48 to 72 hours after injection. The value of cathartics to rid the bowel of normal gallium activity is questionable. 4,5 At our institution, they are not routinely used. To obtain diagnostically useful images, a multipeak gamma camera equipped with a suitable medium energy collimator is mandatory. To facilitate a review of the current role of gallium imaging in infection, patients have been arbitraily divided into three broad categories that are based on the indications most often given for requesting the study: (1) those with fever of unknown origin (FUO); (2) those with suspected osteomyelitis; and (3) those with impaired immunocompetence.
occasions and no diagnosis is established after at least 1 week of inpatient evaluation. 6 Although these are well defined, specific criteria, a review of the imaging literature suggests a more liberal use of the term FUO, which includes febrile patients in whom no definitive diagnosis has been established without regard to the duration or degree of fever or length of hospital stay. Only approximately 25% of all FUOs are caused by infection, and for a technique to be useful, it must be specific; this is particularly true for the abdomen, where nearly one third of the infections responsible for FUOs are located. Although rapid whole-body surveys can be performed with both 67Ga and labeled leukocyte imaging, the advantages of the white-cell technique (described above) make it the current radionuclide procedure of choice for imaging the patient with an FUO with an overall accuracy ranging from 85% to 100%. 7-1~ Nevertheless there is a role for gallium imaging in this population. Sfakianakis et al n prospectively compared gallium imaging with leukocyte imaging for diagnosing occult sepsis. Although the overall accuracy of the two techniques (gallium, 90%; labeled leukocyte, 94%) was nearly identical, the investigators found white-cell imaging to be most useful in patients who were symptomatic for less than 2 weeks, whereas gallium imaging performed better in patients who were symptomatic for more than 2 weeks. They concluded that a negative leukocyte study in a patient symptomatic for more than 2 weeks should be followed by gallium scintigraphy (Fig 2). Under certain circumstances, the nonspecificity of gallium imaging actually may be advantageous. Although a negative white-cell study excludes infection as the cause of an FUO, it does not identify the cause of the fever. A gallium scan performed under these circumstances can, when positive, direct the clinician to a fertile site for additional investigation and may even identify an occult neoplasm. For this reason, for those patients with an FUO and a negative white-cell study, regardless of the duration of symptoms, we routinely recommend gallium scintigraphy (Fig 3).
FUO The strict definition of an FUO is an illness of not less than 3 weeks' duration, during which time a patient's fever exceeds 38.3~ on several
OSTEOMYELITIS
Three-phase bone scintigraphy using [99mTC] MDP is the radionuclide procedure of choice
GALLIUM AND INFECTION
131
Fig 2. Indium-Ill-labeled white cell (left) and STGa (right) images performed on a patient with a 1-month history of fever, elevated white count, and left lower-quadrant pain. The white-cell study is normal, whereas a linear focus of activity extends obliquely through the left lower quadrant on the gallium image (arrow). A psoas abscess was surgically drained.
for diagnosing uncomplicated osteomyelitis, with an accuracy exceeding 90%. 12Abnormalities on bone scintigraphy reflect increased bone-mineral turnover, not infection per se, and this technique has been less successful in diagnosing osteomyelitis superimposed on noninfectious conditions that cause increased bone-mineral turnover. Under these conditions, the specificity of the study has ranged from 0% to 78%. 12 Gallium scintigraphy has been used in conjunction with bone scintigraphy in an effort to enhance the specificity of the technique. Lisbona et a113found that sequential bone/gallium imaging was superior to bone scintigraphy alone for distinguishing cellulitis from osteomyelitis, for precise localization of a focus of infection, and for separating active (acute) from inactive (chronic) osteomyelitis. RosenthaU et aP 4 used combined bone/gallium scintigraphy to evaluate suspected osteomyelitis in patients with orthopedic hardware. They found that infection was not likely to be present when the gallium images were negative (regardless of whether the bone images were positive or negative) or when the distribution of activity was similar on both bone and gallium images (congruent images). They also noted that infection was likely to be present when images were spatially congruent and the uptake of gallium was intense or when the distribution of the two radiotracers was different (spatially incongruent images) (Fig 4). Schauwecker et al, 15 as part of a larger study,
evaluated combined bone/gallium scintigraphy for diagnosing complicating osteomyelitis. Although they found that this technique was excellent for excluding osteomyelitis when the study was normal, and for diagnosing osteomyelitis when the intensity of uptake of 67Ga exceeded the uptake of [99mTc]MDP or when the images were spatially incongruent, these criteria were met in only 28% of the patients studied. These investigators concluded that 67Ga images added little diagnostic information not available on bone scintigraphy alone. In this series, 111In-granulocyte imaging, combined with [99mTc]MDP imaging, was superior to bone/ gallium scintigraphy for diagnosing acute osteomyelitis. Merkel et a116 reviewed sequential bone/ gallium imaging performed on 130 patients with painful orthopedic prostheses. Images were interpreted using criteria similar to those of Rosenthall et al, except that Merkel et al added an equivocal category that included studies in which the distribution and intensity of uptake were similar (or congruent) on both bone and gallium images. These investigators reported an overall accuracy of 77% for the sequential technique. Gomez-Luzuriaga et a117 compared sequential bone/gallium imaging with 111In-leukocytes in 40 patients with suspected total hip replacement infection. The accuracy of the bone/ gallium technique in their series was 80% versus 93% for white-cell imaging. Similar results for
132
CHRISTOPHER J. PALESTRO
Fig 3. Anterior chest image from a ["-Tc]HMPAO-labeled leukocyte study (A) performed on a 9-year-old boy with a I-week history of fever was normal (except for splenomegaly). 67GalIJum-67 chest image (B) shows increased activity in the left upper thorax, Gallium-67 coronal SPECT images (C) confirm that the abnormality extends into the left axilla. Lymph node biopsy showed Hodgkin's lymphoma.
combined bone/gallium imaging have been reported by other investigators, with the overall accuracy of the technique ranging between 60% and 80%. ls-2~ Although initial results using labeled leukocyte imaging for diagnosing complicating osteomyelitis were variable, recent data indicate that when leukocyte imaging is combined with marrow imaging, the diagnostic accuracy of the study consistently exceeds 90%, 2123 making this dual-modality technique the procedure of choice for diagnosing complicated osteomyelitis. The results of white-cell imaging are less than
satisfactory, however, for diagnosing vertebral osteomyelitis. We have reported that the technique is neither sensitive nor specific, with an overall accuracy of approximately 66%. 24In contrast, Modic et a125 evaluated sequential bone/ gallium imaging and reported an accuracy of 86% for diagnosing vertebral osteomyelitis, similar to the accuracy they found for magnetic resonance imaging (MRI). Combined bone/gallium imaging recently has been shown to be useful in both tuberculous and nontuberculous infectious spondylitis. 26 Although the data are limited, they do suggest that bone/gallium imaging may
GALLIUM AND INFECTION
133
Fig 4. Technetium-99-MDP (A) and S~Ga(B) forearm images from an adolescent patient with fractures of the left radius and ulna and the right fourth metacarpal. Bone image shows diffusely increased uptake throughout the left ulna and radius as well as the right fourth metacarpal. Gallium uptake is confined to the left ulnar diaphysis (incongruent distribution) and is intense, consistent with osteomyelitis, which was operatively confirmed. There is virtually no gallium uptake in the left radius, which excludes osteomyelitis. The distribution of the two radiotracers in the right fourth metacarpal is spatially congruent, but the intensity of gallium uptake is much less than that of MDP, and the study therefore is negative for osteomyelitis.
be superior to labeled leukocyte imaging for vertebral osteomyelitis; at the very least, this technique clearly merits further investigation. THE IMMUNOCOMPROMISED PATIENT
The immunocompromised individual has an increased susceptibility to numerous infections that often are caused by organisms not typically pathogenic in the immunocompetent host. These infections can involve virtually every organ system in the body, and despite the often serious nature of these ailments, patients frequently present with minimal signs and symptoms. The ability to perform whole-body surveys rapidly and in a timely fashion, combined with reasonable accuracy, makes radionuclide techniques a particularly attractive diagnostic tool in this population. The discussion that follows focuses on radionuclide imaging of the chest and abdomen, and although based on results obtained in acquired immunodeficiency disease (AIDS) patients, it is applicable to the general population
of immunocompromised people, including those who are receiving antineoplastic therapy and the ever increasing number of organ-transplant recipients. The Thorax
Radionuclide techniques have not figured prominently in the diagnosis of respiratory infection in the general population, probably be, cause sputum cultures and chest radiographs are easily obtainable and usually diagnostic, In contrast to their limited role in the general population, radionuclide studies, notably gallium, have come to play a crucial role in the diagnosis of AIDS-related respiratory disease. Four distinct patterns of gallium uptake in the thorax have been observed, and each pattern is associated with a particular entity or group of entities. The four patterns are (1) normal, (2) lymph-node uptake, (3) focal pulmonary parenchymal uptake, and (4) diffuse pulmonary parenchymal uptake.
134
Normal The normal gallium scan of the chest, particularly in the setting 0 f a normal chest x-ray, excludes infection with a high degree of certainty. Barron et a127 reported that only 1 of 15 patients with negative gallium scans had pulmonary infection, making the predictive value of a negative study 93% in the series. Their 1 false-negative study occurred in a patient with Pneumocystis carinii pneumonia (PCP) who had received 2 weeks of treatment before imaging. None of the eleven patients whose gallium scans and chest x-rays were both negative had pulmonary disease (negative predictive value, 100%). Kramer et a128 reviewed 86 gallium scans performed on 71 human immunodeficiency virus (HIV)-positive individuals and found no evidence of thoracic infection in any of the 20 patients whose gallium studies and chest radiographs both were negative. Woolfenden et a129reported a negative predictive value of 91% when only pulmonary activity greater than marrow activity was the criterion for a positive study. Although the combination of a negative gallium scan and a negative chest x-ray rule out any pulmonary disease with a high degree of certainty, a negative gallium scan in the setting of an abnormal chest x-ray has a different, although no less important, connotation. Among the neoplasms associated with AIDS is Kaposi's sarcoma, which is not gallium avid. In patients with pulmonary Kaposi's sarcoma, a positive gallium scan suggests the presence of superimposed inflammation or infection. A negative scan with a positive chest x-ray indicates that Kaposi's sarcoma is the most likely cause of the patient's deteriorating respiratory status, z8,29 Despite providing strong evidence against pulmonary infection, the negative gallium image occasionally has been observed in patients with clinically obvious progression of respiratory infection, and in this setting, the negative results portend a grave prognosis. Bitran et al 3~ described four such patients, all of whom died within several weeks after undergoing gallium imaging. At autopsy, all were found to have multiple opportunistic infections, including cytomegalovirus, PCP, and Mycobacterium avium intracellulare (MAI). The inflammatory response in each of the patients was abnormal and was characterized by a lack of granulocytes. The
CHRISTOPHER J. PALESTRO
investigators speculated that the absence of galIium Uptake in the Setting Of disseminated infection may have been caused by the aberrant inflammatory response. Lymph node uptake Lymph node uptake of gallium is a frequent finding in the HIV-positive patient (Fig 5). Although such uptake can be identified virtually anywhere in the body, it is most often encountered in the hilar, mediastinal, and para-aortic lymph nodes. Uptake in supraclavicular and cervical lymph nodes often can be seen on gallium images of the chest. Kramer et a128 identified nodal uptake in 10 (12%) of 86 scans performed on HIV-positive patients; this was associated with MAI in 9 patients and with lymphoma in 1 patient. In a later series, 31these same investigators identified nodal uptake in 58 (24%) of 237 scans. MAI or M tuberculosis was present in 35 of the 58 patients, and lymphoma was present in 7. Other conditions associated with nodal uptake of gallium included cryptococcal infection, lymphadenitis, and herpes simplex. Eight (14%) of the 58 patients with nodal uptake had PCP, and all had increased pulmonary parenchymal gallium uptake in addition to nodal uptake. Of particular importance is the fact that in this series, nodal uptake alone was never associated with PCP.
Fig 5. Anterior gallium chest image shows mediastinel, left hilar, and left supraclavicular lymph node uptake as well as faint focal uptake in the right upper-lung parenchyma (arrow) in a patient with MAI pneumonia.
GALLIUM AND INFECTION
Nodal uptake of gallium in the generalized lymphadenopathy associated with the HIVpositive population is variable. Bitran et aP ~ identified lymph node uptake of gallium in 13 (41%) of 32 HIV-positive patients with generalized lymphadenopathy and AIDS-related complex (ARC). Other investigators, however, have not found this entity to be particularly gallium avid. Kramer et aP 2 reported that in their experience, the generalized lymphadenopathy found in patients with ARC is not gallium avid. More recently, Podzamczer et aP 3 reviewed lymph node uptake of gallium in 30 HIVpositive patients with lymphadenopathy. Tissue diagnoses were available in all 30 patients. Lymph node uptake at least as intense as liver uptake was present in 17 patients and was associated with M tuberculosis in 11 patients, MAI in 1 patient, lymphoma in 3 patients, Kaposi's sarcoma with Castelman's disease in 1 patient, and follicular hyperplasia in 1 patient. Three patients had nodal uptake less intense than hepatic uptake, and all had follicular hyperplasia. Ten patients had no gallium uptake in lymph nodes, 9 had follicular hyperplasia, and 1 had Kaposi's sarcoma. Localized (focal) pulmonary uptake This pattern of gallium uptake most frequently is found in the setting of bacterial pneumonia (Fig 6). When the uptake is multifocal, aggressive infections such as actinomycosis or nocardia should be considered. Although gallium studies are often positive in AIDS patients with bacterial pneumonia, corresponding radiographic abnor-
Fig 6. Anterior and posterior eTGa images show focal parenchymal uptake in the right lower lung in a patient with S aureus pneumonia. Note the intense gastric uptake (arrows) in this patient without evidence of gastrointestinal disease,
135
malities invariably are present, and some investigators think that the gallium findings contrib ute little to the diagnosis. 31,34 It is interesting and important to note that other than bacterial pneumonia, the entity most frequently associated with localized pulmonary uptake of gallium is PCP. 31 Diffuse pulmonary uptake. In an effort to maintain consistency in interpretation, facilitate comparison of serial studies, and enhance the diagnostic accuracy of the study, a grading system for evaluating the intensity of pulmonary parenchymal uptake of gallium has evolved. Although the particulars of any given grading system vary somewhat from institution to institution, the intensity of pulmonary gallium uptake commonly is divided into one of four or five grades based on a comparison with some reference standard, which typically includes surrounding soft tissues and the liver. The grading system employed at our institution is as follows (Fig 7): grade 0 (normal), pulmonary uptake is indistinguishable from surrounding soft tissues; grade 1 (equivocal), pulmonary uptake is more intense than grade 0 but less intense than rib activity and is of questionable significance; grade 2 (abnormal), pulmonary uptake is clearly increased but less intense than hepatic activity; a photopenic cardiac silhouette is present; grade 3 (abnormal), pulmonary uptake is increased and equal to hepatic uptake; a photopenic cardiac silhouette is present; and grade 4 (abnormal), pulmonary uptake exceeds liver uptake; a photopenic cardiac silhouette is present.
136
CHRISTOPHER J. PALESTRO
+ i
Fig 7. (A) Grade 0 (normal) gallium image of the chest; (B) Grade 2 diffuse pulmonary uptake; (C) Grade 3 diffuse pulmonary uptake; (D) Grade 4 diffuse pulmonaryuptake {see text for description of grading system).
Diffuse pulmonary uptake of gallium in the AIDS patient is most often associated with PCP, and the overall sensitivity of the technique has generally been reported to exceed 90%. 27-30 Unfortunately, there is a rather long differential diagnosis that accompanies the finding of diffuse pulmonary uptake of gallium, and when this is the sole criterion used for diagnosing PCP, the specificity of the study can be quite low, having been reported to range between 20% and 82%. 31,35 The specificity of the study can be enhanced
by applying additional criteria to the interpretation of the study. Barron et a127 found that a positive gallium scan was more specific for PCP when the chest x-ray was negative than when the x-ray was positive (85% v 74%). Intense pulmonary uptake on gallium images is more likely to be associated with PCP than with other conditions in the AIDS population. Coleman et a135 found that the specificity of the study improved from 20%, when any pulmonary uptake was considered positive for PCP, to 90%, when only pulmonary uptake at least as intense
GALLIUM AND INFECTION
as liver uptake was considered positive. Kramer et a131reported that the specificity of the study increased from 79%, when abnormal lung activity more intense than bone marrow activity was considered positive for PCP, to 95%, when only pulmonary activity more intense than liver activity was considered positive. These investigators also observed that consideration of the distribution of gallium uptake could improve the specificity of the scan for PCP. When diffuse lung uptake was considered positive for PCP, the positive predictive value was 72%; the positive predictive value increasedto 87% when heterogeneous diffuse uptake was the diagnostic criterion for a positive study. Treatment can affect both the intensity and pattern of pulmonary uptake of gallium in AIDS patients. Woolfenden et a129 noted that only 1 of 12 patients with PCP who showed pulmonary uptake at least as intense as liver uptake was being treated at the time of imaging. In contrast, 4 of 7 patients with PCP whose gallium studies showed lung activity less intense than that of the liver were being treated at the time of imaging. Aerosolized pentamidine frequently is used for prophylaxis of recurrent PCP. Gallium uptake, confined to the upper lungs, has been observed in a patient who was treated prophylactically with aerosolized pentamidine and who developed recurrent PCP that was confined to the upper lungs) 6 The proposed mechanism for the regional recurrence of this disease is decreased aerosol deposition in the upper-lung fields. Finally, there are some data that suggest that gallium imaging may be useful to monitor patient response to therapy for PCP. 37,38Touazon et a137 performed gallium imaging on eight patients with PCP subsequently confirmed by fiberoptic bronchoscopy. All eight patients had abnormal gallium scans, including seven patients who had diffuse lung uptake of radiotracer and one patient who had localized uptake. Repeat studies were obtained after treatment in six of the eight patients and were positive in four, three of whom had bronchial washings persistently positive for PCP. The fourth patient subsequently received diagnosing of Coccidioides immitus infection. Both patients with negative follow-up gallium scans had negative bronchial washings.
137
In summary, gallium imaging plays an important role in the diagnosis of AIDS-related respiratory illnesses. A negative scan is strong evidence against an infectious process. Lymphnode uptake most often is associated with mycobacterial disease or lymphoma, whereas focal lung parenchymal uptake usually is seen in the setting of bacterial infection. Although it may be because of a myriad of causes, diffuse lung uptake should be considered suspicious for PCP, especially when the uptake is intense or heterogeneous (Fig 8). The Abdomen The gastrointestinal (GI) tract is the largest lymphoid organ in the human body, and consequently, it is a potential reservoir for the HIV. The GI tract is an important site for HIVinduced immunodeficiency and is predisposed to a broad spectrum of protozoal, fungal, viral, and bacterial diseases that result in considerable morbidity and a not insignificant mortality. 39 Infectious agents that frequently involve the upper GI tract include Candida albicans, herpes simplex virus, cytomegalovirus, and Mycobacterium avium intracellulare. Those organisms with a predilection for the bowel include cryptosporidium, microsporidium, isospora belli, salmonella, clostridium difficile, and herpes simplex. Clinically, patients with AIDS-related illnesses of the upper GI tract, especially the esophagus, present with dysphagia, whereas those with intestinal disorders usually present with abdominal pain, cramping, and diarrhea. Despite the fact that nearly 90% of AIDS patients are affected by HIV-related GI disease at some time during the course of their illness, radionuclide imaging has played a limited role in the diagnostic evaluation of these entities. Whereas there are undoubtedly many reasons for this, two factors are particularly important. Successful treatment of AIDS-related GI disease invariably requires precise identification of the causative agent. In contrast to the thorax, where patterns and intensity of radiogallium uptake often can correctly characterize the identity of a pathogen, no such correlates exist in the abdomen. Moreover, normal excretion of this radiotracer via the large bowel further complicates matters by making differentiation of normal from abnormal bowel activity difficult (Fig 9).
138
CHRISTOPHER J. PALESTRO
'TGa Image
I
I
I
I
I
Normal
Focal Nodal Uptake
Focal Parenchymal Uptake I
Diffuse Parenchymal Uptake
I CXR(-)
No pulmonary disease
I CXR(+)
Kaposi's sarcoma
I
I
I
M. tuberculosis MAI
Bacterial pneumonia
Mild Homogeneous
Intense I
Lymphoma
PCP
PCP TreatedPCP CMV MAI Interstitial pneumonitis
I
I CXR(-)
I Heterogeneous
1 PCP
Fig 8. Outline of patterns of thoracic uptake of eTGa and the entities most often associated with these patterns in AIDS patients.
Fig 9. Intense colonic gallium activity in two AIDS patients. (A) Image is from a patient with pseudomembranous colitis. (B) Image is from a person with no gastrointestinal disease.
GALLIUM AND INFECTION
Thete are scant data available pertaining to gallium scintigraphy of HIV-related illnesses !nvolving the esophagus and stomach. Although a case of gallium uptake in esophageal candidiasis has been observed, the investigators reviewed 100 additional gallium scans, performed primarily on immunocompromised patients, and did not detect any additional cases, noting that a major limitation to gallium imaging of this entity is the presence of sternal and vertebral activity. 4~ Gallium accumulation in gastric lymphoma of an AIDS patient has been described41; yet, gastric uptake, in the absence of any pathology, also has been reported in this population, and the presence of gastric activity therefore must be interpreted with caution (Fig 6). 28,42 In the intestinal tract itself, although gallium accumulation in HIV-related small bowel disease has been reported, 32 the vast majority of the data available pertain to gallium scintigraphy of HIV colitides. Woo!fenden et a129 reviewed large-bowel uptake of radiogallium in 95 AIDS patients (164 scans). Images were performed 48 to 72 hours after injection of 3.5 mCi (130 MBq) (3.5 mCi) of 67Ga. Cathartics were not routinely used. The intensity of colonic uptake was graded on a scale of 0 to 5, with 0 being indistinguishable from background and 5 greater being than liver. Thirty-six patients (38%) showed colonic activity more intense than hepatic uptake (grade 5) in at least one segment. Twenty-three of the 36 patients (64%) were symptomatic with either abdominal pain or diarrhea or both. Fourteen of the 36 patients had colonoscopy or microbiological studies. Eight of the 14 (57%) were diagnosed with GI infection. No definite evidence of infection could be found in the remaining 6 patients, all of whom had normal endosc0pies. Three of these 6 individuals underwent colonic biopsy, and all had evidence of chronic inflammation, Of the 59 patients with bowel uptake less intense than hepatic activity, 4 (7%) had GI infections. These investigators concluded that although the finding of colonic activity on gallium images should be interpreted cautiously, unusually intense (large bowel) activity should be considered suspicious for infection. Tatsch et a143 evaluated the reliability of gallium scintigraphy for detection of AIDSrelated intestinal infections in 25 HIV-positive patients. Imaging was performed at 48 and 72
139
hours after injection of 5 mCi (185 MBq) of 67Ga, and cathartics were routinely used before imaging. Studies were considered abnormal only when bowel activity was unchanged between 48 and 72 hours images (regardless of intensity). Seventeen of the 25 patients (68%) had abnormal gallium accumulation; in 14 the abnormalities involved either all or part of the colon, and in the remaining 3 the abnormalities were localized to abdominal lymph nodes. All 12 of the 17 patients with abnormal gallium uptake who underwent diagnostic evaluation had confirmed AIDS-related diseases; 10 patients with abnormal colonic activity had GI infection, and 2 patients with para-aortic lymph node uptake had atypical mycobacterial infection (Fig 10). The 5 remaining patients all had abnormal colonic activity; 2 patients had diarrhea and negative stool cultures and underwent no further GI workup. Three patients were totally asymptomatic and underwent no GI workup whatsoever. No information was provided about the 8 patients with negative abdominal gallium studies. Slizofski et a144 retrospectively reviewed 21 67Ga studies performed on 19 AIDS patients. Patients received 10 mCi (370 MBq) of 67Ga and were usually imaged at 48 to 72 hours after injection. The investigators did not indicate whether or not cathartics were used. Fourteen of the 19 patients underwent more than one set of images. Colonic activity was classified as abnormal when it was more intense than hepatic activity and did not move on successive scans. Seven patients had acute GI disease; gallium scans were true positive in 4 patients and false negative in 3 patients. Among the 14 studies performed on patients without acute GI disease, there were 10 true negatives and 4 false positives. Not surprisingly, the results of gallium imaging in the abdomen are more variable than those in the chest. Although absence of any colonic activity probably excludes significant bowel pathology, the presence of such activity, even when very intense, must be interpreted cautiously. There are few data available comparing gallium with labeled leukocyte imaging in the immunocompromised population. Based on a comparison of these two techniques in 36 febrile AIDS patients, Fineman et a145 concluded that
140
CHRISTOPHER J. PALESTRO
O Fig 10. (A) 67Ga uptake in celiac and para-aortic lymph nodes in an AIDS patient with MAI. (B) Extensive, intense lymph-node uptake of eTGa in celiac, paraaortie, iliac, inguinal and femoral lymph nodes in an HIV-positive patient with lymphoma.
although labeled leukocyte imaging was superior to gallium imaging for the detection of sinusitis, bacterial pneumonia, and bowel infection, gallium imaging was the procedure of choice for detecting lymph node abnormalities and the opportunistic respiratory infections so prevalent in this population. The role of gallium imaging in infection has
changed dramatically over the past decade, not only because of the development of newer radionuclide techniques, but also because of the ever-increasing number of patients with altered immunocompetence. Rather than resulting in the elimination of gallium imaging for infection, these events have simply redefined its indications.
REFERENCES 1. Edwards CL, Hayes RL: Tumor scanning with 67Ga citrate. J Nucl Med 10:103-105, 1969 2. Lavender JP, Lowe J, Barker JR, et al: Gallium-67 citrate scanning in neoplastic and inflammatory lesions. Br J Radio144:361-366, 1971 3. Hoffer P: Gallium: Mechanisms. J Nucl Med 21:282285, 1980 4. Zeman RK, Ryerson TW: The value of bowel preparation in Ga-67 citrate scanning: concise communication. J Nucl Med 18:886-889, 1977 5. Silberstein EB, Fernandez-Ulloa M, Hall J: Are oral cathartics of value in optimizing the gallium scan? Concise communication. J Nucl Med 22:424-427, 1981 6. Petersdorf RG, Beeson PB: Fever of unexplained origin: Report on 100 cases. Medicine 40:1-30, 1961 7. McDougall IR, Baumert JE, Lantieri RL: Evaluation of 111In leukocyte whole body scanning. AJR 133:849-854, 1979
8. Schmidt KG, Rasmussen JW, Sorensen PG, et al: Indium-lll-granulocyte scintigraphy in the evaluation of patients with fever of undetermined origin. Scand J Infect Dis 19:339-345, 1987 9. Syrjala MT, Valtonen V, Liewendahl K, et al: Diagnostic significance of indium-Ill granul0cyte scintigraphy in febrile patients. J Nucl Med 28:155-160, 1987 10. Kelly MJ, KalffV, Hicks RJ, et al: 11qn-oxine labelled leukocyte scintigraphy in the detection and localization of active inflammation and sepsis. Med J Aust 152:352-357, 1990 11. Sfakinakis GN, AI-Sheikh W, Heal A, et al: Comparisons of scintigraphy with In-111 leukocytes and Ga-67 in the diagnosis of occult sepsis. J Nucl Med 23:618-626, 1982 12. Schauwecker DS: The scintigraphic diagnosis of osteomyelitis. AJR 158:9-18, 1992 13. Lisbona R, Rosenthall L: Observations on the sequential use of 99mTc-phosphate complex and 67Ga imaging in
GALLIUM AND INFECTION
osteomyelitis, cellulitis, and septic arthritis. Radiology 123: 123-129, 1977 14. Rosenthall L, Lisbona R, Hernandez M, et al: 99mTcPP and 67Ga imaging following insertion of orthopedic devices. Radiology 133:717-721, 1979 15. Schauwecker DS, Park HM. Mock BH. et al: Evaluation of complicating osteomyelitis with Tc-99m MDP, In111 granulocytes, and Ga-67 citrate. J Nucl Med 25:849-853, 1984 16. Merkel KD, Brown ML. Fitzgerald RH Jr: Sequential technetium-99m HMDP-gallium-67 citrate imaging for the evaluation of infection in the painful prosthesis. J Nucl Med 27:1413-1417. 1986 17. Gomez-Luzuriaga MA, Galan V, Villar JM: Scintigraphy with Tc. Ga and In in painful total hip prostheses. Int Orthop 12:163-167. 1988 18. Merkel KD, Brown ML, Dewanjee MK, el al: Comparison of indium-labeled-leukocyte imaging with sequential technetium-gallium scanning in the diagnosis of lowgrade musculoskeletal sepsis. J Bone Joint Surg 67:465-476. 1985 19. Merkel KD, Fitzgerald RH Jr. Brown ML: Scintigraphic examination of total hip arthroplasty: Comparison of indium with technetium-gallium in the loose and infected canine arthroplasty, in Welch RB. (ed): The Hip. Proceedings of the Twelfth Open Scientific Meeting of the Hip Society, Atlanta, GA, 1984, pp 163-192 20. AI-Sheikh W. Sfakianakis GN, Mnaymneh W, et al: Subacute and chronic bone infections: Diagnosis using I n - l l l , Ga-67 and Tc-99m MDP bone scintigraphy, and radiography. Radiology 155:501-506, 1985 21. Palestro CJ, Kim CK, Swyer AJ. et al: Total hip arthroplasty: Periprosthetic 111In labeled leukocyte activity and complementary 99mTc sulfur colloid imaging in suspected infection. J Nucl Med 31:1950-1955, 1990 22. Palestro CJ, Swyer AJ, Kim CK, Goldsmith SJ: Infected knee prosthesis: Diagnosis with In-Ill-leukocyte, Tc-99m sulfur colloid, and Tc-99m MDP imaging. Radiology 179:645-648, 1991 23. Palestro CJ, Roumanas P. Swyer AJ, et al: Diagnosis of musculoskeletal infection using combined In-111 labeled leukocyte and Tc-99m SC marrow imaging. Clin Nucl Med 17:269-273, 1992 24. Palestro CJ, Kim CK, Swyer AJ, et al: Radionuclide diagnosis of vertebral osteomyelitis: Indium-Ill-leukocyte and technetium-99m-methylene diphosphonate bone scintigraphy. J Nucl Med 32:1861-1865, 1991 25. Modic MT, Feiglin DH, Piraino DW, et al: Vertebral osteomyelitis: Assessment using MR. Radiology 157:157166, 1985 26. Lisbona R, Derbekyan V, Novales-Diaz J, et al: Gallium-67 scintigraphy in tuberculous and nontuberculous infectious spondylitis. J Nucl Med 34:853-859, 1993 27. Barron TF, Birnbaum NS, Shane LB, et al: Pneumocystis carinii pneumonia studied by gaUium-67 scanning. Radiology 154:791-793, 1985 28. Kramer EL, Sanger J J, Garay SM, et al: Gallium-67 scans of the chest in patients with acquired immunodeficiency syndrome. J Nucl Med 28:1107-1114, 1987 29. Woolfenden JM, Carrasquil!o JA, Larson SM, et al: Acquired immunodeficiency syndrome: Ga-67 citrate imaging. Radiology 162:383-387, 1987
141
30. Bitran J, Bekerman C, Weinstein R, et al: Patterns of gallium-67 scintigraphy in patients with Acquired Immunodeficiency Syndrome and the AIDS related complex. J Nucl Med 28:1103-1106, 1987 31. Kramer EL, Sanger JH, Garay SM, et al: Diagnostic implications of Ga-67 chest scan patterns in human immunodeficiency virus-seropositive patients. Radiology 170:671676, 1989 32. Kramer EL, Sanger J J: Nuclear Medicine in the management of the AIDS patient, in Freeman LM (ed): Nuclear Medicine Annual, New York. NY, Raven. pp 37-57, 1990 33. Podzamczer D. Ricart I, Bolao F, et al: Gallium-67 scan for distinguishing follicular hyperplasia from other AIDS-associated diseases in lymph nodes. AIDS 4:683-685. 1990 34. Kramer EL. Sanger JJ: Detection of thoracic infections by nuclear medicine techniques in the Acquired Immunodeficiency Syndrome. Radiol Clin North Am 27: 1067-1076. 1989 35. Coleman DL, Hattner RS. Luce JM, et al: Correlation between gallium lung scans and fiberoptic bronchoscopy in patients with suspected pneumocystis carinii pneumonia, and the acquired immunodeficiency syndrome. Am Rev Respir Dis 130:1166-1169, 1984 36. q-upler RH, Turbiner EH: Scintigraphic distribution of Pneumocystis carinii pneumonia in AIDS treated with prophylactic inhaled pentamidine. Clin Nucl Med 16:772. 1991 37. Tuazon CU, Delaney MD. Simon GL, et al: Utility of gallium67scintigraphy and bronchial washings in the diagnosis and treatment of Pneumocystis carinli pneumonia in patients with the acquired immune deficiency syndrome. Am Rev Respir Dis 132:1087-1092, 1985 38. Moser E, Tatsch K, Kirsch C-M, et al: Value of 67Gallium scintigraphy in primary diagnosis and follow-up of opportunistic pneumonia in patients with AIDS. Lung 168: 692-703, 1990 (suppl) 39. Smith PD, Quinn TC, Strober W, et al: Gastrointestinal infections in AIDS (NIH Conference). Ann Int Med 116:63-77, 1992 40. Rundback JH, Goldfarb CR, Ongseng F: Gallium-67 imaging in candidal esophagitis. Clin Nucl Med 15:38-39, 1990 41. Gianfelice D, Rosenthall L, Falutz J: Gallium-67 detection of occult gastric lymphoma in AIDS. A JR 149:305306, 1987 42. McNeill JA, Llaurado JG: Innocent intramural gastric uptake of gallium-67 in a case of AIDS. Clin Nucl Med 1!:123-126, 1986 43. Tatsch K, Knesewitsch P, Matuschke A, et al: 67Gascintigraphy for evaluation of AIDS-related intestinal infections. Nucl Med Comm 11:649-655, 1990 44. Slizofski WJ, Brown SJ, Dadparvar S, et al: Diagnostic significance of apparent abnormal bowel activity on Ga-67 scans in AIDS patients. Clin Nucl Med 16:473-477, 1991 45. Fineman DS, Palestro CJ, Kim CK, et al: Detection of abnormalities in febrile AID S patients with In-Illlabeled leukocyte and Ga-67 scintigraphy. Radiology 170: 677-680, 1989