22.2.1 History of Present Illness

A 43-year-old woman who initially presented approximately 1 year after noticing a small nodule over the right superolateral globe which eventually was associated with mild right temporal discomfort. The condition was initially treated as a prolapsed lacrimal gland that was repositioned surgically. Biopsy at the time demonstrated mild chronic inflammation. Follow-up imaging performed a year later showed progressive enlargement. Excisional biopsy of the nodule demonstrated ACC. Follow-up imaging noted persistent asymmetry of the right lacrimal gland but no abnormally enhancing lesion.

Past medical history: Left arm and ankle fracture after car accident.

Past surgical history: Laparoscopic cholecystectomy 2011, nasal septum repair 1988.

Family history: Father died of stomach cancer; mother alive at age 66 with chronic obstructive pulmonary disease (COPD).

Brother and sister in their 40 s are in good health.

Social history: University professor. Denies smoking or other chemical exposures. Social drinker.

Review of systems: Some blurry vision and eye pain.

Neurological examination: CN II-IX, XI, XII intact, PERRL, EOMI, 5/5 motor strength and sensation.

Imaging studies: See Fig. 22.1 and Fig. 22.2.

22.2.2 Treatment Plan

There is a significant reduction in cause-specific mortality at 5 years with intra-arterial cytoreductive chemotherapy (IACC). Patient was admitted for neoadjuvant IAC with cisplatin and Adriamycin on day 1 followed by IACC with Adriamycin for another 2 days. She was premedicated with decadron, Zofran, and Ativan, and started on Neulasta on treatment day 2.

22.2.3 Follow-up

Patient initially complained of right-sided tinnitus and tingling of the right head and extremities which dissipated by day 3. She had no hematologic alterations, cardiovascular changes, or nephrotoxicity. On her last clinic visit 5 years posttreatment, she is doing well. She complains of occasional dry eyes but denies blurry vision, headache, or local symptomatology.

22.3.1 Adenoid Cystic Carcinoma of the Lacrimal Gland

Adenoid cystic carcinoma (ACC) is a rare tumor of the oral and maxillofacial regions predominately found in the minor salivary glands, ¹ but can occur in any secretory gland including the breast, cervix, esophagus, lungs, prostate, and lacrimal gland. ² Lacrimal gland ACC carries a poor prognosis despite disfiguring wide local excision because of its propensity for perineural spread and distant temporal recurrence. ³ The current standard of care is radical orbitectomy, but eye-sparing procedures have been gaining support in the literature because of similar survival rates to radical surgical excision. ⁴

Recurrence most commonly occurs locally. ⁵ For this reason, adjuvant radiation therapy and chemotherapy on top of surgical excision is considered the most promising treatment option. ⁶ A case report ⁷ and larger case series ⁸ have reported improved overall survival and decreased recurrence with IAC for lacrimal gland ACC followed by orbital exenteration, chemoradiation, and IV chemotherapy. Patients who completed treatment had a significantly higher 10-year disease-free survival when compared to the group undergoing conventional treatment (100 vs. 14.3%); however, this group excluded those who were not given IAC due to sacrifice of the lacrimal artery from excisional biopsy, precluding straightforward catheterization. IAC resulted in downstaging primary tumors with intracranial extension or temporal fossa involvement to disease limited to the intraorbital compartment. One major adverse event occurred in a patient who was treated through the ICA and sustained central retinal artery occlusion and complete vision loss prior to exenteration.

A smaller case series of four patients who underwent IAC followed by resection reported significant adverse effects including eye lid and globe necrosis, facial palsy, difficulty chewing, facial swelling, and neutropenia. ⁹ Three patients had total exenteration, and at follow-up (range 48-69 mo), there was no evidence of recurrence of metastases. The fourth patient had lacrimal gland resection alone and at 54 months was found to have local recurrence for which total exenteration was recommended.

Stage I and II disease or tumors less than 4 cm without ex-traparenchymal extension do well with globe preservation surgery and radiation, with local recurrence observed in those with greater residual tumor burden. ¹⁰ Given the high morbidity of orbital exenteration, globe-sparing surgery is gaining popularity with adjuvant proton radiation therapy or concurrent chemoradiation. ¹¹ IAC may play a growing role in reducing microscopic tumor burden.

22.3.2 Intracranial Gliomas

For newly diagnosed malignant gliomas, the standard of treatment entails surgical resection with adjuvant chemoradiation, ^{12 , 13 , 14} yielding an overall survival of 15 months. ^{15 , 16} Methylated 0[6]-methylguanine-DNA methyltransferase (MGMT)-positive histology confers a better prognosis with overall survival of 23.4 months. ¹⁷ IAC is currently utilized either in clinical trials or as off-label salvage therapy. ¹⁸

There are four phase III randomized controlled trials (RCTs) comparing IA with systemic chemotherapy ^{19 , 20 , 21 , 22} in patients with newly diagnosed glioblastoma multiforme. All patients underwent surgical resection followed by radiation therapy with IA carmustine (BCNU), ^{19 , 21} nimustine (ACNU), ²⁰ and carbo-platin and ACNU. ²² These initial chemotherapeutic agents were chosen for their lipophilic properties. ²³ IA therapy was administered through the internal carotid and vertebral arteries, and in some cases the dose was divided between arteries, if the tumor crossed arterial territories. ²¹ Three of the four studies were terminated at interim analysis. The trials’ cumulative 469 patients successfully randomized showed no survival benefit for IAC.

Thousands of patients have been enrolled in phase I and II clinical trials testing the safety and efficacy of IAC for newly diagnosed as well as recurrent grade III to IV gliomas. Early small case series for newly diagnosed malignant gliomas focused on nitrosoureas (ACNU, BCNU, and hydroxyethyl-chloroethyl nitrosourea [HeCNU]) on the basis that lipophilic agents would facilitate passage across the blood-brain barrier. Leukoencephal-opathy and ocular toxicity were common side effects, prompting combination therapy. Median survival ranges from 5 to 22 months (Table 22.1, ▶ Table 22.2).

With recurrent tumors, overall survival is 25 weeks. ²⁴ One multi-institutional RCT found a statistically significant survival benefit of IV PCNU over IA cisplatin (13 vs. 10 mo). ²⁵ Although IV PCNU was associated with greater hematotoxicity, IA cisplatin led to greater renal toxicity, ototoxicity, and encephalopathy.

Similar to the literature on newly diagnosed malignant gliomas, early trials in the treatment of recurrent tumors ( Table 22.3) initially focused on nitrosoureas before moving toward platinum-based agents (cisplatin, carboplatin). Reports of ocular, hematologic, and neurologic toxicities likely spurred interest in immunologic agents and combination IA, IV, and PO therapy. In the recurrent tumor IA literature, there are more reports of these new agents as well as the use of hydrophilic compounds with adjuvant blood-brain barrier disruption (BBBD). ²⁶ Despite greater patient volume and outcomes data, conclusions are difficult to make on the whole because of treatment and patient heterogeneity. Many studies include not only patients with various prior surgical, medical, and radiation histories but also add adjuvant or concurrent radiation in the treatment arm along with multiagent therapy (▶ Table 22.4).

Poor support to IA therapy may be related to the heterogeneity of the underlying pathology, toxicity of earlier chemotherapeutic agents, and poor patient selection. MGMT gene expression confers a particular benefit from temozolomide and may be used for patient selection for super selective intra-arterial chemotherapy (SSIAC) trials. ^{27 , 28} Other prognostic factors for survival include young age, good clinical performance status, extensive surgical resection, seizures, and histologic features. ²⁹ In the highly selected populations undergoing IAC, it is difficult to surmise which patient characteristics predict a good response. In one series of 41 patients receiving IAC, 80% of which were glioblastomas (GBMs), half of the tumors were avascular and the other half very hypervascular on angiography. Hypervascularity was a statistically significant predictor of decreased survival. ³⁰ Imaging related predictors of increased survival included small enhancing tumor volume, avascularity, and longer time to recurrence which was defined as clinical deterioration or more than 25% increase in tumor area on magnetic resonance imaging (MRI) (fluid-attenuated inversion recovery [FLAIR] hyperintensity or enhancement). Small tumor volume and avascularity are thought to be related as more aggressive tumors tend to more vascular because of increased micro-vessel density ³¹ and tumoral angiogenesis. ³²

Major hurdles in selective drug delivery include the an-gioarchitecture of GBMs which derive vascular supply from adjacent vascular territories or even the contralateral hemisphere. ²³ Postcontrast enhancement on MRI typically underestimates the vascular territory involved, a finding overcompen-sated by T2 signal abnormalities. ³³ Using FLAIR hyperintensity, 66% of tumors were supplied by both anterior and posterior territories and 10% were supplied by three arterial territories. This suggests that catheterization of a single intracranial artery is insufficient, and effective treatment likely involves superse-lection of branches in different vascular territories.

Most recent SSIAC case series have been performed in those with recurrent high-grade gliomas. ^{34 , 35 , 36} One group has published phase I and II clinical trials employing SSIAC with bevaci-zumab for recurrent glioblastoma. ^{34 , 35} IAC was administered with concomitant balloon occlusion of downstream parent vessels. Follow-up MRI at 1 month also demonstrated decrease in tumor volume as measured by FLAIR hyperintensity, perfusion volume, and enhancement. IAC conferred a median progression-free (PFS) and overall (MOS) survival of 10 and 8.8 months, respectively, which was comparable to prior studies using IV bevacizumab where medial PFS and MOS ranged 3.7-4.2 and 7.2-9.2 months, respectively. ^{37 , 38}

Qureshi et al reported a case series of 12 GBM patients treated with SSIAC carboplatin and nonapeptide H-Arg-Pro-Hyp-Gly-Thi-Ser-Pro-Tyr(Me)-ip(CH2NH)-Arg-OH (RMP-7) for BBBD (discussed later). All were cases of recurrent GBM after gross total resection and chemoradiation. In the six patients with GBM for whom follow-up data was available, one-half demonstrated decreased tumor volume on MRI, defined as enhancement on Tl-weighted images, at a median follow-up of 4.9 months. One-year survival for GBM patients was 45% (5 out of ll). ³⁶

SSIAC decreases nontarget drug delivery while increasing the tumor tissue dose, particularly those supplied by small distal branches. Superselective catheterization of the middle cerebral artery (MCA) was associated with fewer ocular toxicities. ³⁹ In later studies employing SSIAC, there is a notable decrease in the incidence of seizures, leukoencephalopathy, and hematologic toxicities. ^{34 , 35 , 36} Interestingly, those receiving intraprocedural RMP-7 for BBBD had fewer neurological complications, though not statistically significant. ³⁶

The advantage of SSIAC is direct administration to the tumor bed to achieve high local concentrations. To decrease streaming, pulsatile dosing and spatial dose fractionation was used to deliver a more tailored dose for the territory supplied by a given artery. This accounts for individual variations in the circle of Willis so that dosing is based on arterial territory volume and not body weight or body surface area as a way to decrease neurologic complications from administering chemotherapy to normal tissues. ¹⁸

Advances in the understanding of drug delivery have informed clinical trials. Initially postcontrast enhancement on MRI of the brain was used to indicate blood-brain barrier breakdown and support the use of IAC. ⁴⁰ While degree of enhancement does reflect tissue uptake of chemotherapeutic drugs, ²⁶ animal studies comparing MRI enhancement with drug uptake in tumor versus normal tissues have demonstrated unacceptably high doses in normal brain tissue as well as systemically. ⁴¹ This has fueled research in local disruption of the blood-brain barrier to facilitate drug delivery.

Clinical trials employing IAC with BBBD are accruing patients and results are promising. The most commonly employed agent is hyperosmolar mannitol infused right before chemotherapy administration to induce osmotic shrinkage of the endothelial cells lining the capillaries to promote separation of tight junctions. ⁴² An alternative agent is lobradimil (also known as RMP-7 or Cereport), a bradykinin derivative that increases vascular permeability, and regadenoson an adenosine agonist causing transient BBBD. ⁴³

In addition to superselective catheterization, other strategies for efficient drug delivery include flow arrest to maximize drug extraction and minimize drug streaming. High cerebral blood flow and low drug concentrations hinder tissue uptake. Streaming can be overcome with injections of drug volumes greater than 20% of the background blood flow rate, injecting during diastole, or using catheters with side ports. ^{44 , 45 , 46} These delivery pharmacokinetics are studied using the Dedrick model of blood flow. ⁴⁷

Convection-enhanced drug delivery is another catheter-based strategy where catheters are directly implanted into the tumor bed. Drugs are pressure infused at a set concentration, rate, and duration across the leaky cytoarchitecture of growing tumor cells ⁴⁸ and across pre-existing tracts of white matter edema. ⁴⁹ Small phase I and II trials have experimented with monoclonal antibodies, liposomal vectors packaged for gene therapy, antisense oligonucleotides, and exotoxins. ⁵⁰ However, early phase III trials were not able to demonstrate a survival benefit, ⁵¹ although PFS was significantly increased in the treatment arm, 17.7 versus 11.4 weeks.

Riina et al have reported the use of SSIAC with bevacizumab and mannitol for a malignant glioma in 2009 and for a malignant brainstem glioma in 2010 with decreased enhancement seen on follow-up MRI in both cases. ^{52 , 53} Numerous clinical trials have also described IAC in the treatment of diffuse intrinsic pontine gliomas (DIPGs), ⁵⁴ esthesioneuroblastoma, ⁵⁵ anaplastic oligodendrogliomas, ⁵⁶ germ cell tumors, ⁵⁷ cerebral metastatic disease. However, these treatments are entirely experimental.

This limited availability has led to international controversy in the media for rare malignancies for which there is no accepted treatment. ⁵⁸